achievement gaps: how hispanic and white students in ... · achievement gap between hispanic and...

Achievement GapsHow Hispanic and White Students in Public Schools Perform in Mathematics and Reading on the National Assessment of Educational ProgressStatistical Analysis Report

NCES 2011-459 U.S. DEPARTMENT OF EDUCATION

Achievement GapsHow Hispanic and White Students in Public Schools Perform in Mathematics and Reading on the National Assessment of Educational Progress

Statistical Analysis Report

June 2011

F. Cadelle Hemphill Alan VannemanNAEP Education Statistics Services Institute

Taslima Rahman Project OfficerNational Center for Education Statistics

U.S. Department of EducationNCES 2011-459

U.S. Department of Education Arne Duncan Secretary

Institute of Education Sciences John Q. Easton Director

National Center for Education Statistics Jack Buckley Commissioner

The National Center for Education Statistics (NCES) is the primary federal entity for collecting, analyzing, and reporting data related to education in the United States and other nations. It fulfills a congressional mandate to collect, collate, analyze, and report full and com-plete statistics on the condition of education in the United States; conduct and publish reports and specialized analyses of the meaning and significance of such statistics; assist state and local education agencies in improving their statistical systems; and review and report on education activities in foreign countries.

NCES activities are designed to address high-priority education data needs; provide consistent, reliable, complete, and accurate indica-tors of education status and trends; and report timely, useful, and high-quality data to the U.S. Department of Education, the Congress, the states, other education policymakers, practitioners, data users, and the general public. Unless specifically noted, all information con-tained herein is in the public domain.

We strive to make our products available in a variety of formats and in language that is appropriate to a variety of audiences. You, as our customer, are the best judge of our success in communicating information effectively. If you have any comments or suggestions about this or any other NCES product or report, we would like to hear from you. Please direct your comments to

National Center for Education Statistics Institute of Education Sciences U.S. Department of Education 1990 K Street NW Washington, DC 20006-5651

June 2011

The NCES Home Page address is http://nces.ed.gov.The NCES Publications and Products address is http://nces.ed.gov/pubsearch.

This report was prepared for the National Center for Education Statistics under Contract No. ED-05-R0014 with the American Institutes for Research.

Suggested CitationHemphill, F.C., and Vanneman, A. (2011). AchievementGaps:HowHispanicandWhiteStudentsinPublicSchoolsPerforminMathematicsandReadingontheNationalAssessmentofEducationalProgress (NCES 2011-459). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, DC.

For ordering information on this report, write to

ED Pubs U.S. Department of Education P.O. Box 22207 Alexandria, VA 22304

or call toll free 1-877-4ED-Pubs or order online at http://www.edpubs.gov.

Content Contact Taslima Rahman (202) 502-7316 [email protected]

The National Assessment of Educational Progress (NAEP), a congressionally mandated project of the U.S. Department of Education, informs the public periodically about the academic achievement of elementary and secondary students in reading, mathematics, science, writing and other subjects. Only information related to academic achievement and relevant variables is collected under this program from students representing the country. By making objective information available on performance of all race/ethnic groups at the national and state levels, NAEP is an integral part of our nation’s evaluation of the condition and progress of education. While the National Center for Education Statistics (NCES) within the Institute of Education Sciences (IES) of the U.S. Department of Education conducts the survey, the National Assessment Governing Board oversees and sets policies for NAEP.

iii

Executive SummaryThis report provides detailed information on the size of the achievement gaps between Hispanic and White public school students at the national and state levels and describes how those achievement gaps have changed over time. Additional information about race/ethnicity in NAEP is given in appendix A. Most of the data in this report is derived from the results of the 2009 National Assessment of Educational Progress (NAEP) main assess-ments in mathematics and reading; however the trend data provided is derived from results from as early as 1990. AchievementGaps:HowHispanicandWhiteStudentsin Public Schools Perform in Mathematics and Reading onthe National Assessment of Educational Progress, follows our previous report that provided similar information on the achievement gap between Black and White students (Vanneman et al. 2009).

Hispanics are the fastest-growing segment of the United States population. According to the U.S. Census Bureau data (Guzman 2001), the Hispanic1 population increased by about 58 percent, from 22 million in 1990 to 35 million in 2000, compared with an increase of about 13 percent for the total U.S. population. In 2010, the U.S. Census Bureau estimated the number of Hispanics to be about 50.5 million, or about 16 percent of the U.S. population, up 43 percent from the 2000 census. The increase of over 15 million Hispanics from 2000 to 2010 accounted for more than half of the total popu-lation increase in the U.S. during that time (Humes, Jones, and Ramirez 2011). As these data reflect, the proportion of the U.S. population that is Hispanic is increasing over time. Additionally, data collected in 2009 by the U.S. Department of Education indicate that a substantial proportion of Hispanic students in grades 4 (37 percent) and 8 (21 percent) are English language learners (table 2). These two facts—the growing size of the Hispanic population in the United States and the percentage of fourth- and eighth-grade Hispanic students that are English language learners—underlie the achievement gap between Hispanic and White fourth- and

1 According to the U.S. Census, Hispanics or Latinos are those people who clas-sified themselves in one of the specific Spanish, Hispanic, or Latino categories listed on the Census 2010 questionnaire. People who identify their origin as Spanish, Hispanic, or Latino may be of any race. For further information see U.S. Census Bureau, 2010 Census of Population, Public Law 94-171 Redistricting Data File. Available online: http://factfinder2.census.gov.

eighth-graders. Closing the Hispanic-White achievement gap remains a challenge. While Hispanic students’ average scores have increased across the assessment years, White students had higher scores, on average, on all assessments.

The NAEP 2009 Reading and Mathematics Assessments included grade 4 and grade 8 students nationally and for all 50 states, as well as the District of Columbia and the Department of Defense Education Activity (hereinafter referred to as states).2

MathematicsIn 2009, NAEP mathematics scores for both Hispanic and White students in grades 4 and 8 nationwide were higher than in 1990, the first assessment year for both Hispanic and

2 Not all states had Hispanic (or White) student populations large enough to pro-vide reliable data, and not all states participated in the earliest NAEP assessments.

Table A. Trends in NAEP mathematics at grades 4 and 8 since earliest comparison year, by grade and student group: 2009

Scores Gap Hispanic White

4th Grade

National Public

Gender

Male

Female

NSLP1

Eligible

Not Eligible

8th Grade

National Public

Gender

Male

Female

NSLP1

Eligible Narrowed

Not Eligible no significant change in score gap.

increased score.1 National School Lunch ProgramNOTE: Comparison year for National Public and Gender is 1990; NSLP comparisons are made to 2003.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), Various years: 1990-2009 Mathematics Assessments.

iv

White public school students. Mathematics scores increased, but the achievement gap between Hispanic and White students did not change significantly at either grade 4 or 8 from 1990 to 2009. From 2007 to 2009, scores for Hispanic and White fourth-graders remained unchanged and the gap persisted at 21 points. For eighth-graders, scores increased for both Hispanic and White students from 2007 to 2009, but the gap remained at 26 points, which was not significantly different from the gap in 1990 or 2007. At grade 8, the 2009 mathematics achievement gap for Hispanic and White stu-dents eligible for the National School Lunch Program was narrower than in 2003 (table A).

n In 2009 at grade 4, eleven states had a smaller Hispanic-White gap than the nation, and six states had a gap that was larger (table B).

Table B. State gaps in mathematics compared to the nation: 2009

National Hispanic-White

gap

States with gaps that are:

Smaller than nation Larger than nation

Grade 4 21 points AK, DoDEA,1 FL, GA, KY, LA, MO, MT,

NY, OK, WY

CA, CT, DC, MA, RI, UT

Grade 8 26 points AK, AR, DE, DoDEA,1 FL, GA, HI, IN, KY, MI, MO, OK,

TN, VA, WY

CA, CO, CT, NY, RI, WA

1 Department of Defense Education Activity (overseas and domestic schools).NOTE: Gaps are significantly different (p<.05) from the national gap.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Mathematics Assessment.

Table C. State trends in mathematics score gaps: Various years, 1990–2009

Since first assessment Since 2007

Grade 4

Narrowed CT, DC, DE, DoDEA,1 MA, MI, MO, NJ, NY, OR, RI

None

Widened None RI, TX

Grade 8

Narrowed CT, DE, HI, MO, RI AR, DE

Widened MD, UT None1 Department of Defense Education Activity (overseas and domestic schools).SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1990-2009 Mathematics Assessments.

n At grade 4, in all 21 states for which 1992 data were available, both Hispanic and White students achieved higher average scores in mathematics in 2009 than in 1992. In six of those states (Connecticut, the District of Columbia, Massachusetts, New Jersey, New York, and Rhode Island) the gap narrowed as Hispanic students’ scores increased more than White students’ scores. In five additional states (Delaware, the Department of Defense Education Activity, Michigan, Missouri, and Oregon) the gap narrowed between Hispanic and White students since the first NAEP assessment year for that state or the first year for which Hispanic student results are reportable. Since all states did not participate in the grade 4 NAEP mathematics assessment in 1992, the first NAEP assessment year varies (table C).

n In 2009 at grade 8, fifteen states had a smaller Hispanic-White gap than the nation, and six had a gap that was larger (table B).

n At grade 8, in 14 of the 15 states for which 1990 data were available, the mathematics scores of Hispanic and White students were higher in 2009 than in 1990. In both Connecticut and Rhode Island, the gap was nar-rower in 2009 than in 1990. In three additional states, Delaware, Hawaii, and Missouri, the gap narrowed between Hispanic and White students since the first year for which Hispanic student results are reportable.

n In Maryland, the gap was wider in 2009 than in 1990, as White eighth-graders’ scores increased more than those of their Hispanic peers. In Utah, the gap was wider in 2009 than in 1992, the first NAEP assessment year for that state. Since all states did not participate in the grade 8 NAEP mathematics assessment in 1990, the first NAEP assessment year varies (table C).

n Hispanic-White mathematics gap data were not avail-able in 2009 for fourth graders in Maine, Mississippi, North Dakota, Vermont, or West Virginia, or for eighth-graders in the District of Columbia, Louisiana, Maine, Mississippi, North Dakota, Vermont, or West Virginia because the size of the NAEP sample of Hispanic or White students was too small to provide reliable results.

v

ReadingAt the national level, reading scores increased for both groups significantly, but the achievement gap between Hispanic and White students did not change for fourth- or eighth-graders when comparing 1992 to 2009. From 2007 to 2009, scores did not change significantly for either group at the fourth grade. The 26-point gap for fourth-graders in 2007 was not significantly different from the 25-point gap in 2009. The 25-point gap for eighth-graders in 2007 was not significantly different from the 24-point gap in 2009, though scores for both Hispanic and White students have increased. At grades 4 and 8, the 2009 reading achieve-ment gap for Hispanic and White students eligible for the National School Lunch Program was narrower than in 2003 (table D).

n At grade 4, thirteen states had a smaller Hispanic-

Table D. Trends in NAEP reading at grades 4 and 8 since earliest comparison year, by grade and student group: 2009

Scores Gap Hispanic White

4th Grade

National Public

Gender

Male

Female

NSLP1

Eligible Narrowed

Not Eligible

8th Grade

National Public

Gender

Male

Female

NSLP1

Eligible Narrowed

Not Eligible no significant change in score or score gap.

increased score.1 National School Lunch ProgramNOTE: Comparison year for National Public and Gender is 1990; NSLP comparisons are made to 2003.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), Various years: 1992-2009 Reading Assessments.

White gap than the nation, and six had a gap that was larger (table E).

n At grade 4, in 11 of the 21 states for which 1992 data were available, the reading scores of Hispanic and White stu-dents were higher in 2009 than in 1992. Both New Jersey and New York had a narrower gap in 2009 than 1992. In Colorado, the gap widened when comparing 2009 to 1992. In Indiana the gap widened between Hispanic and White students when comparing 2009 to 2002, the first NAEP assessment year for which Hispanic student results are reportable for that state. All states did not participate in the first grade 4 state NAEP reading assessment in 1992, so the first year for which data were available varies (table F).

n At grade 8, seven states had a smaller Hispanic-White gap than the nation, and no state had a gap that was larger (table E).

Table E. State gaps in reading compared to the nation: 2009

National Hispanic-White

gap

States with gaps that are:

Smaller than nation Larger than nation

Grade 4 25 points AK, DE, DoDEA,1 FL, HI, IA, KY, LA, MD,

MO, MT, SD, WY

CA, CO, CT, DC, MN, UT

Grade 8 24 points AK, DoDEA,1 FL, KY, MO, SC, WY

None

1 Department of Defense Education Activity (overseas and domestic schools).NOTE: Gaps are significantly different (p<.05) from the national gap.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Reading Assessment.

Table F. State trends in reading score gaps: Various years, 1992–2009

Since first assessment Since 2007

Grade 4

Narrowed NJ, NY AK

Widened CO, IN None

Grade 8

Narrowed AK RI, SC, WY

Widened None NoneSOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1992-2009 Reading Assessments.

vi

n At grade 8 state-level data were available for 22 states starting in 1998. When comparing 2009 to 1998, the grade 8 reading gap did not change significantly in any state. In Wyoming, both Hispanic and White students scored higher in 2009 than in 1998. In Alaska, the gap narrowed between Hispanic and White students when comparing 2009 to 2003, the first NAEP assessment year for that state. All states did not participate in the first grade 8 state NAEP reading assessment in 1998, so the first year for which data were available varies (table F ).

n Hispanic-White reading gap data were not available in 2009 for fourth-graders in Maine, North Dakota, Vermont, or West Virginia, or for eighth-graders in the District of Columbia, Louisiana, Maine, Mississippi, Montana, North Dakota, South Dakota, Vermont, or West Virginia because the size of the NAEP sample of Hispanic and White students was too small to provide reliable results.

The NAEP reading and mathematics scales make it possible to examine relationships between students’ per-formance and various background factors measured by NAEP, such as race. However, a relationship that exists between achievement and another variable does not reveal its underlying cause, which may be influenced by a num-ber of other variables. Similarly, the assessments do not reflect the influence of unmeasured variables. The results are most useful when they are considered in combination with other information about the student population and the education system, such as trends in instruction, changes in the school-age population, and societal demands and expectations.

All differences discussed in this report are significant at the .05 level after controlling for multiple comparisons. The technical notes for this report provide information about sampling, accommodations, interpreting statistical signifi-cance, and other technical features.

ix

ContentsExecutive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1The distribution of the Hispanic population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3The data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Understanding score gaps and the ways gaps can change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5Understanding statistical significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6Cautions in interpreting the data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7How this report is organized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

NAEP Mathematics

National Results for Hispanic and White Fourth- and Eighth-Graders . . . . . . . . . . . . . . . . . . . . . . . . . 10Mathematics scores and achievement gaps in the nation, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Mathematics scores and achievement gaps by gender, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Mathematics scores and gaps by family income, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Trends in mathematics scores and achievement gaps by ELL status, 1996–2009 . . . . . . . . . . . . . . . . . . . . . . 16

State Results for Hispanic and White Fourth- and Eighth-Graders . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18State and national mathematics achievement gaps at grade 4, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18Trends in state mathematics achievement gaps at grade 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20State and national mathematics achievement gaps at grade 8, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Trends in state mathematics achievement gaps at grade 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Mathematics Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34National trends in mathematics, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34State achievement gaps in mathematics, compared to the nation, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34State trends in mathematics, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

NAEP Reading

National Results for Hispanic and White Fourth- and Eighth-Graders . . . . . . . . . . . . . . . . . . . . . . . . . 36Reading scores and achievement gaps in the nation, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Reading scores and achievement gaps by gender, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Reading scores and gaps by family income, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Trends in reading scores and achievement gaps by ELL status at grades 4 and 8, 1998–2009 . . . . . . . . . . . . . 42

State Results for Hispanic and White Fourth- and Eighth-Graders . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44State and national reading achievement gaps at grade 4, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44Trends in state reading achievement gaps at grade 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46State and national reading achievement gaps at grade 8, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52Trends in state reading achievement gaps at grade 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Reading Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60National trends in reading, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60State achievement gaps in reading, compared to the nation, 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60State trends in reading, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

x

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

Appendix A: Technical Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64Frameworks, development, administration, scoring, and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64Sources of the data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64NAEP sampling procedures for public school students in reading and mathematics . . . . . . . . . . . . . . . . . . . . . 64School and student participation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Understanding NAEP reporting groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Race/ethnicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68National School Lunch Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Inclusion and exclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Accommodations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70Drawing inferences from the results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70Weighting and variance estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72Analyzing group differences in averages and percentages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72Conducting multiple tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74Cautions in interpretation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Appendix B: Supplemental Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

xi

List of TablesTable Page

A. Trends in NAEP mathematics at grades 4 and 8 since earliest comparison year, by grade and student group: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

B. State gaps in mathematics compared to the nation: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

C. State trends in mathematics score gaps: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

D. Trends in NAEP reading at grades 4 and 8 since earliest comparison year, by grade and student group: 2009 . . v

E. State gaps in reading compared to the nation: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

F. State trends in reading score gaps: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

1. Percentage of public school students assessed in NAEP mathematics eligible for the National School Lunch Program, by race/ethnicity and grade: Various years, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . .14

2. Percentage of public school English language learner students assessed in NAEP mathematics, by race/ethnicity and grade: Various years, 1996–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

3. Percentage of public school students assessed in NAEP reading eligible for the National School Lunch Program, by race/ethnicity and grade: Various years, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .40

4. Percentage of public school English language learner students assessed in NAEP reading, by race/ethnicity and grade: Various years, 1998–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42

A-1. School and student participation rates, and target student population, grade 8 reading assessment, public school students only, by state or jurisdiction: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66

A-2. National mathematics and reading exclusion rates as percentages of the total sample, public schools only, by grade and race/ethnicity: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70

A-3. Mathematics and reading exclusion rates as percentages of the total sample, public schools only, by grade, race/ethnicity, and jurisdiction: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71

B-1. Administration of NAEP national and state mathematics assessments, by grade: Various years, 1990–2009 . .76

B-2. Average national mathematics scale scores for all public school students at grades 4 and 8, by gender and eligibility for the National School Lunch Program: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . .76

B-3. Administration of NAEP national and state reading assessments, by grade: Various years, 1992–2009 . . . . .77

B-4. Average national mathematics scale scores for all public school students at grades 4 and 8, by gender and eligibility for the National School Lunch Program: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . .77

B-5. Percentages of public school students in NAEP mathematics and reading classified as English language learners, by subject, grade, race/ethnicity, and jurisdiction: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78

B-6. Percentages of public school students in NAEP mathematics and reading classified as eligible for a free or reduced-price school lunch, by subject, grade, race/ethnicity, and jurisdiction: 2009 . . . . . . . . . . . . . . . . . .79

xiii

List of FiguresFigure Page

1. Percentage of Hispanic public school students in the NAEP reading assessment at grade 4, by state: 2003 and 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2

2. Hispanic and English language learner population in NAEP reading in selected states at grade 4: 2009 . . . . . . .3

3. Administration of main NAEP national and state mathematics and reading assessments: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

4. Ways gaps can narrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6

5. Mathematics achievement score gaps between Hispanic and White public school students at grade 4: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

6. Mathematics achievement score gaps between Hispanic and White public school students at grade 8: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

7. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by gender: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

8. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by gender: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

9. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by eligibility for the National School Lunch Program: Various years, 2003–2009 . . . . . . . . . . . . . . . . . . . . . .15

10. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by eligibility for the National School Lunch Program: Various years, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . . . 15

11. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by English language learner status: Various years, 1996–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

12. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by English language learner status: Various years, 1996–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

13. The Hispanic-White achievement score gap in mathematics for public school students at grade 4, by state: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19

14. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by state: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

15. The Hispanic-White achievement score gap in mathematics for public school students at grade 8, by state: 2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27

16. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by state: Various years, 1990–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29

17. Reading achievement score gaps between Hispanic and White public school students at grade 4: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

18. Reading achievement score gaps between Hispanic and White public school students at grade 8: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

19. Reading achievement score gaps between Hispanic and White public school students at grade 4, by gender: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39

20. Reading achievement score gaps between Hispanic and White public school students at grade 8, by gender: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39

21. Reading achievement score gaps between Hispanic and White public school students at grade 4, by eligibility for free or reduced-price school lunch: Various years, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . .41

xiv

22. Reading achievement score gaps between Hispanic and White public school students at grade 8, by eligibility for free or reduced-price school lunch: Various years, 2003–2009 . . . . . . . . . . . . . . . . . . . . . . .41

23. Reading achievement score gaps between Hispanic and White public school students at grade 4, by English language learner status: Various years, 1998–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43

24. Reading achievement score gaps between Hispanic and White public school students at grade 8, by English language learner status: Various years, 1998–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43

25. The Hispanic-White achievement score gap in reading for public school students at grade 4, by state: 2009 . .45

26. Reading achievement score gaps between Hispanic and White public school students at grade 4, by state: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47

27. The Hispanic-White achievement score gap in reading for public school students at grade 8, by state: 2009 . .53

28. Reading achievement score gaps between Hispanic and White public school students at grade 8, by state: Various years, 1992–2009 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55

Figure Page

1

IWhen the earliest National Assessment of Educational Progress (NAEP) reading assessment was administered in 1971, no separate scores were recorded for any racial/ethnic groups except White and Black students. Together, these two groups included about 98 percent of all U.S. students surveyed by NAEP (Campbell, Hombo, and Mazzeo 2000). All other students, constituting two percent of the popula-tion, were classified as “Other.” In 2009, at the fourth grade, 56 percent of all U.S. students were White, 16 percent Black, 21 percent Hispanic, 5 percent Asian/Pacific Islander, and 1 percent American Indian/Alaska Native, according to 2009 NAEP Mathematics Assessment data. In less than 40 years, Hispanic students have gone from being an almost unob-served racial/ethnic group to the largest, and fastest growing, racial/ethnic group in the United States (Humes, Jones, and Ramirez 2011). In three states—New Mexico, California and Texas—about 50 percent of the fourth-grade popula-tion is of Hispanic descent. Figure 1 shows the change in the percentage of Hispanic students in grade 4 from 2003 to 2009 based on NAEP reading data.

As measured by NAEP, the educational performance of Hispanic students has generally lagged behind the perfor-mance of White students. The gap in scores between White and Hispanic students in mathematics in 1990, when the current main NAEP mathematics assessment was first administered, was not significantly different from the gap in scores in 2009, for either grade 4 or grade 8 (National Center for Education Statistics 2009). The same is true in reading, comparing results in 2009 with the scores for 1992, when the current main NAEP reading assessment was first admin-istered (National Center for Education Statistics 2010). At the state level, in 2009 gaps between Hispanic and White students were statistically significant in almost every state for which reliable results were available in both reading and mathematics at both grades 4 and 8.

The major questions addressed in this study are:

1) How do score gaps in 2009 mathematics and reading per-formance compare to the gaps in the initial and most recent prior years of the NAEP national and state assessment series?

ntroduction2) How do Hispanic and White scores and gaps in mathematics

and reading at the state level compare to the national scores and gaps in 2009?

The Elementary and Secondary Education Act of 1965 intended to improve the education achievement of low-per-forming students in reading and mathematics. Subsequent reauthorizations of the act have reaffirmed the importance of closing the achievement gaps. This report uses NAEP data to examine the progress of the nation and each of the states in reducing the gap between Hispanic and White students at grades 4 and 8 in both reading and mathemat-ics. Because NAEP is designed to report results for public school students at the state level, all of the results that appear in the body of this report, including national results, are for public school students only.

Issues relating to the Hispanic-White achievement gap have been addressed by a number of recent studies. Status and Trends in the Education of Racial and Ethnic Groups (KewalRamani, Fox, and Aud 2010), issued by the National Center for Education Statistics (NCES), exam-ined the educational performance and attainment of all major racial and ethnic groups in the United States from prekindergarten through the postsecondary level, along with employment and income data for these groups. The report identified a variety of factors that may be associated with the achievement gap between Hispanic and White students. For example, Hispanic students were more likely than White students to come from low-income families (as defined by student eligibility for the National School Lunch Program), which is associated with lower educa-tional performance. Other reports have also used NAEP data, as well as data from other sources, in analyses to identify important factors related to the Hispanic-White achievement gap. Parsing the Achievement Gap II (Barton and Coley 2009) examined educational achievement gaps in terms of differences in life experiences among racial/ethnic and socioeconomic subgroups that could affect differences in academic achievement. The report also reviewed stu-dent performance over time to determine if the gaps were changing. Latinos and Education: Explaining the Attainment Gap (Lopez 2009), a national survey of Latinos, concluded

2

that the biggest reason for the relatively limited educational attainment of many Latinos was due to a decision to discon-tinue education in order to support a fam-ily. Many Latinos also discontinued their education because of poor English skills or a dislike of school, the study said. How Far Behind in Math and Reading are English Language Learners? (Fry 2007) examined the performance of English language learn-ers, using NAEP data supplemented with results from selected state assessments. The Family: America’s Smallest School (Barton and Coley 2007) examined the effects of children’s home life on academic perfor-mance of Hispanic and other students in terms of such factors as out-of-wedlock births, two-parent versus one-parent fami-lies, family income, home literacy develop-ment, child care, educational resources in the home, and the parent-school relation-ship. In many cases, Hispanic children were more likely than White children to be raised in circumstances associated with below average academic performance—lack of two parents in the home, for exam-ple, or low family income, or access to quality day care. Hispanics and the Future of America (National Research Council 2006) analyzed data from a variety of sources to identify differences, if any, between Hispanics and other immigrant and minority groups. The report also discussed the likelihood of social integration of both immigrant and native-born Hispanics.

Intr

oduc

tion

Figure 1. Percentage of Hispanic public school students in the NAEP reading assessment at grade 4, by state: 2003 and 2009

2003

AL1%

AK4%

AZ36%

AR4%

CA47%

CO23%

FL 21%

GA6%

HI3%

ID13%

IL16%

IN5%

IA5%

KS8%

KY1%

LA1%

ME1%

MA11%

MI5%

MN4%

MS1%

MO3%

MT2%

NE9%NV

28%

NH2%

NM51%

NY21%

NC6%

ND2%

OH2%

OK7%

OR14%

PA4%

SC3%

SD2%

TN2%

TX42%

UT11%

VT1%

VA5%

WA12%

WV#

WI6%WY

8%

National Public18%

2009

0–5% Hispanic 6–10% Hispanic 11–15% Hispanic 16–25% Hispanic 26+% Hispanic

National Public21%

CT14%

MD5%

NJ16%

DE8%

RI18%

DC9%

DoDEA14%

AL4%

AK7%

AZ45%

AR8%

CA51%

CO29%

FL 24%

GA10%

HI3%

ID13%

IL21%

IN6%

IA7%

KS14%

KY3%

LA3%

ME1%

MA17%

MI5%

MN7%

MS2%

MO4%

MT3%

NE15%NV

39%

NH3%

NM56%

NY19%

NC10%

ND2%

OH3%

OK9%

OR16%

PA8%

SC5%

SD3%

TN5%

TX49%

UT15%

VT1%

VA8%

WA18%

WV1%

WI9%WY

10% CT16%

MD10%

NJ19%

DE12%

RI18%

DC11%

DoDEA16%

# Rounds to zero.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2003 and 2009 Reading Assessments.

3

The distribution of the Hispanic populationHispanics are the second largest racial/ethnic group in the United States, comprising 16 percent of the nation’s popu-lation in 2010. This was an increase of 43 percent compared to 2000, when Hispanics constituted 12.5 percent of the population (Humes, Jones, and Ramirez 2011).

The concentration of the Hispanic population varies by region. In 2010, the West had a higher percentage of its population who were Hispanic—29 percent. In contrast, in the Midwest the population was 7 percent Hispanic. In the South the percentage was 16 percent; in the Northeast, 13 percent. New Mexico was the state with the largest percentage of its population who were Hispanic (46 percent). Other states with large percent-ages included California and Texas (both 38 percent). California, Florida, Illinois, New York, and Texas are the five states with the largest number of Hispanics (U.S. Census Bureau 2010).

A large percentage of the Hispanic population is foreign born—44 percent in 2007, as compared to 14 percent for the entire U.S. population. In addition, 11 percent of the Hispanic population under 18 was foreign born in 2007. Approximately 65 percent of those who were under 18 and foreign born were Mexican, while about 9 percent were South American, 8 percent were Puerto Rican, 4 percent were Dominican, 3 percent were Salvadoran, 6 percent were Other Central American, and about 3 percent each were Cuban and Other Hispanic/Latino (KewalRamani, Fox, and Aud 2010).

Mexican American students, whether foreign born or native to the United States, made up about two thirds of Hispanic eighth-graders in public schools nationally in 2009, according to data collected by the 2009 NAEP Reading Assessment. They also constituted about 80 per-cent of Hispanic eighth-graders in California, Illinois, and Texas. In Florida, about 18 percent of the Hispanic popula-tion were Mexican American, and in New York about 11 percent. About 23 percent of Hispanic students in Florida were Cuban American.

Figure 2. Hispanic and English language learner population in NAEP reading in selected states at grade 4: 2009

0

20

40

60Percent

51

2924

6

21

7

19

7

49

17

Hispanic ELL Hispanic ELL Hispanic ELL Hispanic ELL Hispanic ELL

California

FloridaIllinois

New York

Texas

NOTE: Other states may have higher Hispanic population density. ELL includes Hispanic and non-Hispanic students.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Reading Assessment.

4

Thirty-five percent of all Hispanic fourth-graders and 20 percent of all Hispanic eighth-graders were identified as English language learners (ELL) in the 2009 NAEP Reading Assessment, compared to 9 percent and 5 percent, respectively, for all students (including Hispanics) at the two grades. Figure 2 shows the percentage of Hispanic and ELL students at grade 4 in the five states with the largest Hispanic population.

In 1998, the first year for which the NAEP Reading Assessment has separate data for ELL Hispanic students, 22 percent of Hispanic students were ELL at grade 4. Fifteen percent at grade 8 were ELL. In 2009, the percent-ages for ELL Hispanic students were 35 percent at grade 4 and 20 percent at grade 8. The differences between the percentages for 1998 and 2009 are only statistically signifi-cant at grade 4.

In this report, the performances of ELL and non-ELL Hispanic students are compared to each other, and the performance of non-ELL Hispanic students is com-pared to that of White students, using national NAEP data (1996–2009 in mathematics, 1998–2009 in reading). Because the percentage of ELL White students is so small (one percent or less), all references to White students will include both ELL and non-ELL White students unless otherwise noted.

The dataThis report compares national public school student per-formance for Hispanic and White students in mathemat-ics and reading for 2009 to their performance in all prior NAEP assessments. NAEP state-level assessments were introduced in different years during the 1990s. Therefore, state-level comparisons go back to 1990 for grade 8 math-ematics, to 1992 for grade 4 reading and mathematics, and to 1998 for grade 8 reading (figure 3).

NAEP assessments allow the examination of trends in the Hispanic and White performance gap in every state, plus the District of Columbia and the Department of Defense

Education Activity (DoDEA) schools. Discussion of NAEP grade 12 assessments is omitted in this report because these assessments were conducted at the national level only prior to 2009. Additional information on the national and state assessments is given in appendix A.

All data presented in this report are for public school stu-dents only. NAEP provides national results for both public and private school students, but NAEP state results are for public school students only. To maintain consistency of data for comparison purposes, this report uses only public school data at the national level as well.

Hispanic-White achievement gaps results for NAEP have been available to users in two ways: 1) online, using the NAEP Data Explorer at http://nces.ed.gov/nationsreportcard/naepdata/, and 2) in print, in the report cards for a given assessment. Achievement Gaps: How Hispanic and White Students in Public Schools Perform in Mathematics and Reading on the National Assessment of Educational Progress is the first NCES publication to present the Hispanic and White NAEP achievement gaps across time for all the states and the nation, including results for every assessment year since state assessments began.

NAEP does not have Hispanic-White gap data for all states going back to the 1990 mathematics and 1992 reading assessments. The 2001 reauthorization of the Elementary and Secondary Education Act required each state to participate in the NAEP mathematics and reading assessments if they were to receive Title I education fund-ing (Public Law 107-110 Title I Part A, Sec. 1111), effective in 2003. Prior to the passage of the Act, participation was voluntary and about 40 states participated in each assess-ment. (In this report, “state” and “jurisdiction” will be used interchangeably to refer to the 50 states, the District of Columbia, and the DoDEA schools.) Beginning in 2003, all 52 states have participated in all NAEP reading and mathematics assessments.

In addition, in many states NAEP did not obtain samples for Hispanic students large enough to permit the reporting

Intr

oduc

tion

http://nces.ed.gov/nationsreportcard/naepdata/

5

Figure 3. Administration of main NAEP national and state mathematics and reading assessments: Various years, 1990–2009

1990 1992 1994 1996 1998 2000 2002 2003 2005 2007 2009

Mathematics

4th GradeNational U U U U U U U U

State U U U U U U U


State U U U U U U U U

Reading

4th GradeNational U U U U U U U U U

State U U U U U U U U


State U U U U U U

SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1990–2009 Assessments.

of reliable results for the early assessments. For example, at grade 4 in mathematics, results for Hispanic students were not reported for 21 states in the state’s first assessment. In the most recent NAEP mathematics assessment in 2009, there were only five states without results for Hispanic students.

Since 1996 (1998 at the state level), students receiving accommodations on their state assessment received the same accommodations on NAEP, except where an accom-modation would change the nature of what is being tested (see appendix A for details). The 2009 mathematics assess-ment results are based on nationally representative samples of 168,800 fourth-graders and 161,700 eighth-graders. The reading assessment results are based on nationally repre-sentative samples of 178,800 fourth-graders and 160,900 eighth-graders. The main NAEP samples are large because they include representative samples for the 52 states. These samples are weighted to compensate for undersampling of the states with large populations and oversampling of the states with small populations, as well as oversampling of

schools with high concentrations of students from certain racial/ethnic groups and the lower sampling rates of stu-dents who attend very small schools.

NAEP assessments are conducted in a six-week window starting in January of each assessment year. Scores for read-ing and mathematics cannot be compared because the two assessments are scaled independently. In addition, com-parisons cannot be made across grades in a single subject, because the fourth- and eighth-grade assessments contain different questions and thus students in the two grades are not taking the same assessment. See appendix A for more details.

Understanding score gaps and the ways gaps can changeThe achievement gap between Hispanic and White students is defined as the difference between the aver-age score for White students and the average score for Hispanic students. Comparisons are made for main

6

NAEP between the most recent assessment year (2009) and all previous assessment years. Only differences between 2009 and the earliest assessment year, and between 2009 and 2007, are discussed in the text, unless a score for 2009 is different from all previous assessment years. The figures indicate all previous assessment scores that differ from 2009.

Changes in the size of the achievement gap depend on changes in the average scores for Hispanic and White students. Generally, increasing scores and narrowing gaps are seen as desirable, while decreasing scores and widen-ing gaps are seen as undesirable. However, it is possible for the Hispanic-White gap to widen when scores for both Hispanic students and White students increase, if scores for White students increase more than scores for Hispanic students. And it is also possible for the gap to narrow when scores for both Hispanic and White students decline, if scores for White students decline more than scores for

Intr

oduc

tion

Figure 4. Ways gaps can narrow

The average scores of both groups increase, while the score of the lower performing group increases even more.

The average score of the higher performing group does not change, while the score of the lower performing group increases.

The average score of the higher performing group declines, while the score of the lower performing group increases.

The average score of the higher performing group declines, while the score of the lower performing group does not change.

The average scores of both groups decline, but the score of the higher performing group declines even more.

The average scores of both groups do not change significantly, but the combined effect causes a significant narrowing of the gap.

Hispanic students. Figure 4 illustrates the various ways that gaps can narrow.

It is important to note that although NAEP data can iden-tify gaps and changes in gaps, these data cannot explain why gaps exist or why they change. NAEP assessments are designed to measure student performance and identify fac-tors associated with it, not to identify or explain the causes of differences in student performance.

Understanding statistical significanceNAEP data are based on samples of students, and the results are subject to sampling and measurement error. Statistical tests are used to determine whether the differences between average scores are statistically significant, that is, whether they exceed the margin of error. It is possible for the size of the achievement gap to increase or decrease even though the average scores of neither Hispanic nor White students changed significantly during the same period.

In several states in 2009, the difference in scores for Hispanic and White students was not statistically signifi-cant—that is, the score difference was not greater than zero—meaning that there was no achievement gap. In some cases, an apparently large difference in one state may not be statistically significant, while an apparently smaller difference in another state may be statistically significant. This is because findings of statistically sig-nificant differences are a function of both the differences in scores and the standard errors associated with those scores. An apparently large score difference may prove not to be statistically significant if the standard errors involved are large as well. The size of the standard error associated with a NAEP scale score is a function, among other things, of the size of the sample and the degree of variability of performance in the sample, as measured by the standard deviation, i.e. the square root of the vari-ance, which is the average of the squared deviations of performances from the mean or average score. The size of the standard errors may also be influenced by other factors, such as how representative the assessed students

7

are of the entire population. For all these reasons, a score difference that is found to be significant in one state may not be found to be significant in another.

The term “significant” is not intended to imply a judg-ment about the absolute magnitude or the educational relevance of the differences. It is intended to identify statistically reliable population differences to help inform discussion among policymakers, educators, researchers, and the public.

Beginning in 2002, the main NAEP national sample was obtained by aggregating the samples from each state, rather than by using an independently selected national sample. As a result, the national samples in mathematics and reading were larger in 2003, 2005, 2007, and 2009 than in previous assessment years. In addition, the percentage of Hispanic students both nationally and in many states has been increasing since the first assessments in 1990 and 1992, tending to increase the size of the sample for Hispanic students in recent years. Thus, smaller score differences between years or between student groups were found to be statistically significant than would have been detected in previous assessments. All differences discussed in the text are significant at the .05 level with appropriate adjustments for part-to-whole and multiple comparisons. See appendix A for more details.

Statistical comparisons of NAEP scores from different assessment years are “pairwise” comparisons, with appro-priate multiple comparison adjustments. In figures 13, 15, 25, and 27, comparisons of the size of the Hispanic-White achievement gap for each state to the national gap are made using pairwise comparisons with part-to-whole adjustments, where each state is compared to the nation one at a time.

Cautions in interpreting the dataAll results given here are in terms of average scores, which reflect a wide range of student performance. Many Hispanic students score above the average for White stu-

dents and many White students score below the average for Hispanic students. For detailed information on varia-tions in performance, including standard deviations, con-sult the NAEP Data Explorer online at http://nces.ed.gov/nationsreportcard/naepdata/.

The analysis of NAEP data contained in this report should not be seen to imply causal relations. Simple cross-tabulations of a variable with measures of educational achievement, like the ones presented here, cannot be con-sidered as evidence that differences in the variable cause differences in educational achievement. As noted earlier, NAEP surveys are not designed to identify causal rela-tionships. There are many possible reasons why the per-formance of one group of students will differ from that of another. Inferences related to student group performance should take into consideration the many socioeconomic and educational factors that may also be associated with performance.

All statistical tests are performed using unrounded scale scores. The Hispanic-White achievement gap is calculated by subtracting the average scale score for Hispanic students from the average scale score for White students. Because all results are presented as rounded numbers, occasionally the lower scale score plus the gap will not equal the higher scale score shown in this report’s graphics.

How this report is organizedThe remainder of this report presents first mathemat-ics and then reading results. The mathematics sec-tion is color-coded with green page margins while the reading section is color-coded with blue margins. In each section, national results appear first. Information on scores and score gaps over time is presented at the national level for fourth- and eighth-grade Hispanic and White public school students. Similar comparisons are included for White and non-ELL Hispanic students and for non-ELL and ELL Hispanic students. National data also include information on scores and score gaps



8

Intr

oduc

tion

over time for Hispanic and White students by gender and by family income as measured by eligibility for the National School Lunch Program (NSLP).

State-level data include scores and score gaps over time for fourth- and eighth-grade Hispanic and White public school students for each state, limited by non-participation of some states in the early NAEP assessments and by the fact that NAEP did not always obtain samples of Hispanic or White students large enough to allow the reporting

of reliable results. Because Hispanic populations have changed significantly over time, percentages of White and Hispanic students are given for each state for the first assessment in which the state participated and for the most recent assessment in 2009. In addition, the size of the gap in 2009 for each state is compared against the Hispanic-White gap nationally and the scores of Hispanic and White students in each state are compared against the national averages for Hispanic and White students.

10

National Results for Hispanic and White Fourth- and Eighth-Graders

Mathematics scores and achievement gaps in the nation, 1990–2009Average fourth-grade mathematics scores for the nation Average mathematics scores were higher in 2009 than in were higher in 2009 than in 1990 for both Hispanic and 1990 for both Hispanic and White eighth-graders (figure White public school students (figure 5). The 21-point 6). The 26-point gap in 2009 was not significantly different gap in 2009 was not significantly different from the from the 24-point gap in 1990. The gap also did not change 19-point gap in 1990. From 2007 to 2009, scores remained from 2007 to 2009. Scores for both groups rose by 2 points, unchanged for both Hispanic and White students and the leaving the gap at 26 points.gap remained at 21 points.

Ma

th

em

at

ics

11

Figure 5. Mathematics achievement score gaps between Hispanic and White public school students at grade 4: Various years, 1990–2009

Scale score

0

175

225

275

500

1990n 1992n 1996 2000 2003 2005 2007 2009

219* 227* 231* 233*243* 246* 248 248

199* 201* 207* 207*221* 225* 227 227

GapWhite

Hispanic19 26* 2421

26*212121

National average

n Accommodations were not permitted for this assessment.

* Significantly different (p<.05) from 2009.NOTE: Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1990–2009 Mathematics Assessments.

Figure 6. Mathematics achievement score gaps between Hispanic and White public school students at grade 8: Various years, 1990–2009

1990n 1992n 1996 2000 2003 2005 2007 2009

Scale score

0

200

250

300

500

269* 276* 279* 283* 287* 288* 290* 292

245* 247* 249* 252* 258* 261* 264* 266

White

HispanicGap

24 29 30 31* 2628* 2626

National average



1iona

l n G

rad

Nat

990 1991 1992 1993 1994 1995 1996 1996 1997 1998 19I

99

II

II

II I I I I

e 4

& 8

9

12

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 19 9 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II

I I I I I I I I I I I I I I I I

Mathematics scores and achievement gaps by gender, 1990–2009Average mathematics scores were higher in 2009 than in In addition to the 1990-2009 gain, eighth-grade mathemat-1990 for the nation’s Hispanic and White students at both ics scores were higher in 2009 than in 2007 for Hispanic fourth and eighth grades, regardless of gender (figures 7 and White male students and for White female students. and 8). However, the Hispanic-White mathematics gap did Neither gender demonstrated a significant change in the not change significantly for either male or female students Hispanic-White mathematics gap from 2007 to 2009, at at either grade when comparing 2009 to 1990. either the fourth or eighth grade.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

13

Figure 7. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by gender: Various years, 1990–2009

Scale score

0

175

225

275

500

1990n 1992n 1996 2000 2003 2005 2007 2009 1990n 1992n 1996 2000 2003 2005 2007 2009

249249247*244*236*232*228*

219*228228227*223*

207*207*199*199*

246246244*241*231*230*225*218*

226226223*220*208*206*203*200*

White

HispanicGap

Male Female

22 212029*

212529*20

2021 21 2124 232218


* Significantly different (p<.05) from 2009. NOTE: Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1990–2009 Mathematics Assessments.

Figure 8. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by gender: Various years, 1990–2009

Scale score

0

225

275

325

500

1990n 1992n 1996 2000 2003 2005 2007 2009 1990n 1992n 1996 2000 2003 2005 2007 2009

291289*287*286*282*279*276*268*

293292*288*287*284*280*275*270*

264264260*257*251*247*246*243*

267265*262*260*252*251*248*246*

White

HispanicGap

Male Female

27 26262829 32*272425 262729*31323125



1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


14

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Mathematics scores and gaps by family income, 2003–2009The National Assessment of Educational Progress (NAEP) uses student eligibility for free or reduced-price school lunch as an indicator of family income. At grade 4, mathematics scores were higher in 2009 than in 2003 for all Hispanic and White public school students, regardless of school-lunch eligibility (figure 9). When comparing 2009 to 2003, the gaps between the scores of Hispanic and White students in 2009 were not significantly different from 2003, regardless of eligibility. The gaps in 2009 were also not significantly different from 2007, regardless of eligibility. Additionally, scores in 2009 for both White and Hispanic students were not significantly different from 2007, regardless of eligibility.

At grade 8, scores in 2009 were higher than in 2003 for Hispanic and White students regardless of eligibility, and higher than in 2007 for not eligible White students (figure 10). For eligible students only, the gap narrowed from 2003 to 2009, declining by 4 points. From 2007 to 2009, there were no significant changes in the gaps.

At grade 4 in 2009, the 11-point achievement gap for eligible students was smaller than the 16-point gap for not eligible students, and both gaps were smaller than the 21-point gap for all grade 4 students (figure 5). The achievement gap is affected by the comparative proportions

Nati

onal

n G

rade

4 &

8

Table 1. Percentage of public school students assessed in NAEP mathematics eligible for the National School Lunch Program, by race/ethnicity and grade: Various years, 2003–2009

Hispanic WhiteGrade 4 2009 77 29 2007 75 27* 2005 77 27* 2003 74* 26*Grade 8 2009 72 25 2007 69* 23* 2005 69* 23* 2003 67* 20*

* Significantly different (p<.05) from 2009.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 2003–2009 Mathematics Assessments.

of Hispanic and White students who are from low-income families. The achievement gap is larger when all students are considered because most Hispanic students (about 77 percent in 2009) come from low-income families—that is, are eligible for free or reduced price lunches—while most White students (70 percent in 2009) come from families with higher incomes and are not eligible (table 1). On average, students from higher income families have higher scores than those from low-income families and the size of the Hispanic-White gap reflects the greater percentage of White students coming from higher-income families.

The same pattern is seen at grade 8. The 13-point gap for eligible students was smaller than the 23-point gap for not eligible students, and both were smaller than the 26-point gap for all grade 8 students. At grade 8, about 72 percent of Hispanic students came from low-income families in 2009, while 74 percent of White students came from higher income families.

NAEP collects data on students’ eligibility for the National School Lunch Program (NSLP)—sometimes referred to as the free or reduced-price school lunch program—as an indicator of family economic status. Eligibility for free or reduced-price lunch is based on students’ family income in relation to the federally established poverty levels.

Not eligible: Students who are not eligible for the pro-gram because their family’s income is above 185 percent of the poverty level.

Eligible: Students who are eligible for either reduced-price lunch because their family’s income is between 130 percent and 185 percent of the poverty level, or for free lunch, because their family’s income is below 130 percent of the poverty level.

As a result of improvements in the quality of the data on students’ eligibility for NSLP, the percentage of students for whom information was not available has decreased in comparison to the percentages reported prior to the 2003 assessment. Therefore, trend comparisons are only made back to 2003 in this report.

Eligibility for free or reduced-price lunch

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Math

ematics

15

It is also possible to compare the performance of eligible and gap between White eligible and White not eligible students. not eligible students by race/ethnicity and then compare the

At grade 8, White not eligible students had an average score size of the gaps. For example, in 2009 at grade 4, White not

of 297 (figure 10, left side of graph), while White eligible stu-eligible students had an average score of 253 (figure 9, left dents had a score of 276 (figure 10 right side of graph), result-side of graph), while White eligible students had a score of ing in a gap of 21 points. Hispanic not eligible students had 236 (figure 9, right side of graph), resulting in a gap of 17

points. Hispanic not eligible students had an average score an average score of 275, while Hispanic eligible students had

of 237, while Hispanic eligible students had a score of 225, a score of 263, resulting in a gap of 12 points, smaller than the

resulting in a gap of 12 points, smaller than the 17-point 21-point gap for White eligible-not eligible students.

Figure 9. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by eligibility for the National School Lunch Program: Various years, 2003–2009

Scale score

0

175

225

275

500

2003 2005 2007 2009 2003 2005 2007 2009

236236234*231*

253252250*247*

225224222*218*237235235232*

WhiteGapHispanic

Not eligible Eligible

1712 12 12 11

15 1615


Figure 10. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by eligibility for the National School Lunch Program: Various years, 2003–2009

Scale score

0

225

275

325

500

2003 2005 2007 2009 2003 2005 2007 2009

276275273*271*

297295*292*291*

263261257*254*

275273271*269*

WhiteGapHispanic

17* 15* 131422 22 2321



1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


16

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Trends in mathematics scores and achievement gaps by ELL status, 1996–2009At grade 4, the gap between White students (both ELL and non-ELL White students) and non-ELL Hispanic students was smaller in 2009, at 14 points, than in the first year, 1996, when it was 20 points (figure 11). From 2007 to 2009, there was no significant change in the scores of either group or in the gap. (For percentages of ELL students in fourth and eighth grade, see table 2.)

Table 2. Percentage of public school English language learner students assessed in NAEP mathematics, by race/ethnicity and grade: Various years, 1996–2009

Hispanic WhiteGrade 4 2009 2007 2005 2003 2000 1996

37 40 39 40 35 31

1111##

Grade 8 2009 2007 2005 2003 2000 1996

21 25 25 25 19 20

#111##

# Rounds to zero.NOTE: Data shown are the percentages of Hispanic students who were ELLs and the percentages of White students who were ELLs. SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1996–2009 Mathematics Assessments.

Nati

onal

n G

rade

4 &

8

Within the Hispanic student population, the 19-point gap between ELL Hispanic and non-ELL Hispanic fourth-graders in 2009 was not significantly different from either comparison year, 1996 or 2007 (figure 11).

At grade 4 in 2009, the 21-point gap between all Hispanic and White students (figure 5) was larger than both the 14-point gap between White and non-ELL Hispanic stu-dents and the 19-point gap between non-ELL Hispanic and ELL Hispanic students (figure 11).

At grade 8, the gap between White students and non-ELL Hispanic students was smaller in 2009, at 19 points, than in 1996 when it was 24 points (figure 12). From 2007 to 2009, there was no significant change in the gap.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

17

Within the Hispanic eighth-grade student population, the 34-point gap between ELL and non-ELL Hispanic students in 2009 was not significantly different from either comparison year.

At grade 8 in 2009, the 34-point gap between ELL and non-ELL Hispanic students was larger than either the 26-point gap for all White and Hispanic students (figure 6)or the 19-point gap for White and non-ELL Hispanic stu-dents. This 19-point gap was the smallest of the three gaps.

Figure 11. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by English language learner status: Various years, 1996–2009

Scale score

0

175

225

275

500

1996 2000 2003 2005 2007 2009 1996 2000 2003 2005 2007 2009

231* 233*243* 246* 248 248

211* 213*228* 232* 235 234 211* 213*

228* 232* 235 234

198* 196*211* 214 214 216

WhiteGapNon-ELL Hispanic

20* 20*1314 13 14

GapELL Hispanic

Non-ELL Hispanic

13 1717 18 20 19

White–Non-ELL Hispanic Non-ELL Hispanic–ELL Hispanic

* Significantly different (p<.05) from 2009.NOTE: Score gaps are calculated based on differences between unrounded average scores. White includes ELL and non-ELL White students.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1996–2009 Mathematics Assessments.

Figure 12. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by English language learner status: Various years, 1996–2009

1996 2000 2003 2005 2007 2009 1996 2000 2003 2005 2007 2009

Scale score

0

200

250

300

500

279* 283* 287* 288* 290* 292

256* 257* 265* 269* 272 273 256* 257* 265* 269* 272 273

223* 229*237* 238 241 239



24* 26* 1921* 18 19

GapELL Hispanic

Non-ELL Hispanic

33 28 29* 30* 31 34

* Significantly different (p<.05) from 2009.NOTE: Score gaps are calculated based on differences between unrounded average scores. White includes ELL and non-ELL White students.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1996–2009 Mathematics Assessments.

18

State Results for Hispanic and White Fourth- and Eighth-Graders

State and national mathematics achievement gaps at grade 4, 2009The NAEP state mathematics assessments were admin-istered to public school eighth-graders only in 1990, and to public school fourth- and eighth-graders in 1992, 1996, 2000, 2003, 2005, 2007, and 2009. Before 2003, states were not required to participate in NAEP to qualify for Title I education funds. Typically, 40 or more states partici-pated in each prior assessment. Since 2003, all 50 states, the District of Columbia, and the DoDEA participated.

State results are presented in two ways. Comparisons of fourth-grade mathematics gaps in 2009 between each state and the nation are presented in figure 13.

Comparisons of the mathematics gaps within a state over time are presented in a series of small graphs in figure 14. At the top left of each two-page spread, the mathematics scores and gaps for the nation are presented for reference. Each state figure, as well as the national figure, also con-tains a dashed gray line representing the national average for public school students. The data for the national aver-ages are located in the appendix in table B-2.

Eleven states had a smaller Hispanic-White gap than the nation’s 21-point gap in 2009 (Arkansas, DoDEA, Florida, Georgia, Kentucky, Louisiana, Missouri, Montana, New York, Oklahoma, and Wyoming) and six had a gap that was larger (California, Connecticut, District of Columbia, Massachusetts, Rhode Island, and Utah). In 30 states, the gap was not significantly different from the nation’s gap (figure 13). Asterisks indicate a significant difference from the national average score or the national average gap.

The fourth-grade mathematics gap in 2009 was statistically significant in 46 of the 47 states for which data could be reported. In Missouri, the 8-point difference between the average scores for Hispanic and White students was not statistically significant, so in that state there was no mea-sureable gap.

For 7 of the 11 states with gaps smaller than the national average (Arkansas, DoDEA, Florida, Georgia, Missouri, Montana, and New York), Hispanic students had an aver-age score that was higher than the national average for

Ma

th

em

at

ics

19

Hispanic students. In Florida, the aver-age score for White students was higher than the national average for White stu-dents as well. In Kentucky, Louisiana, Oklahoma, and Wyoming, the average score for Hispanic students was not mea-surably different from the national average for Hispanic students, while the average score for White students was below the national average for these students.

Among the six states where the Hispanic- White gap was larger than the national average, in Massachusetts scores for both Hispanic and White students were high-er than the national averages for these groups, while in Utah scores for both groups were lower. In California and Rhode Island, scores for Hispanic stu-dents were below the national average, while scores for White students were not measurably different from the national average. In Connecticut and the District of Columbia, scores for Hispanic students were not measurably different from the national average, while scores for White students were above it.

Figure 13. The Hispanic-White achievement score gap in mathematics for public school students at grade 4, by state: 2009

Jurisdictions

200 210 220 230 240 250 260 270 280 5000

Scale score

Wyoming

WashingtonVirginia

UtahTexas

TennesseeSouth Dakota

South CarolinaRhode IslandPennsylvania

OregonOklahoma

OhioNorth Carolina

New YorkNew MexicoNew Jersey

New HampshireNevada

NebraskaMontanaMissouri

MinnesotaMichigan

MassachusettsMarylandLouisianaKentucky

KansasIowa

IndianaIllinoisIdaho

HawaiiGeorgiaFloridaDoDEA1

District of ColumbiaDelaware

ConnecticutColorado

CaliforniaArkansas

ArizonaAlaska

Alabama

National Public 248

249 243*

245* 247

252* 253*

249 270*

245* 250*

247 247

244* 249

247 245*

251* 241* 241*

255* 258*

243* 255*

245* 247

245* 245*

252* 255*

245* 248

254* 249

241* 243*

249 247

245* 247*

239* 254*

246* 251*

247

244*Wisconsin 250*

227

220*232*

220*233*

219*228

227231*

227235*

238*231*230

225227

230223

233* 227

230238*

232*227

232237*

241*224*

227234*

232*224*

231*236*

233229

221*227

219*232233

225233*

219*234*

227

231

237*

228

WhiteHispanic Gap

17 17 23 12* 28* 23 26* 18 43* 10* 12* 15* 17 19 21 16 22 18 14* 10* 17 26* 16 23 8* 6* 21 19 18 24 21 17* 18 16 12* 21 22 28* 13 13 14 20 27* 17 20 22 13*

21

* Significantly different (p<.05) from the nation (public) when comparing one state to the nation at a time.1 Department of Defense Education Activity (overseas and domestic schools).NOTE: Reporting standards not met for Maine, Mississippi, North Dakota, Vermont and West Virginia because their Hispanic student population size was insufficient for comparison. They are not included in the figure. Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Mathematics Assessment.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


20

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Trends in state mathematics achievement gaps at grade 4The Hispanic-White mathematics gap among the nation’s public school fourth-graders was narrower in 2009 than in 1992, as Hispanic students’ scores showed a greater gain than White students’ scores (figure 14, national results).

From 2007 to 2009, scores of Hispanic and White fourth-graders in the nation did not change significantly. Additionally, there was no significant change in the gap.

In all 21 states for which 1992 data were available, both Hispanic students and White students achieved higher aver-age scores in mathematics in 2009 than in 1992. Six of these state —Connecticut, District of Columbia, Massachusetts, New Jersey, New York, and Rhode Island—also narrowed the achievement gap as Hispanic students’ scores increased more than White students’ scores.

In five states—Delaware, DoDEA, Michigan, Missouri, and Oregon—the achievement gap was narrower in 2009 than in the first year for which reportable results were avail-able for both groups of students: Delaware (1996); DoDEA (1996); Michigan (1996); Missouri (2003); Oregon (1996).

From 2007 to 2009, scores for both White and Hispanic students rose in the District of Columbia. However, the Hispanic-White mathematics gap at grade 4 did not decrease significantly in the District of Columbia or any state when comparing 2009 to 2007.

In both Rhode Island and Texas, the Hispanic-White mathematics gap widened between 2007 and 2009. In Rhode Island, scores for White students increased while scores for Hispanic students did not change significantly.

Stat

e n G

rade

4

Changing of the Gap

In the following 11 states, the gap was narrower in 2009 than in the first assessment year for which reli-able results for both groups were available as Hispanic

Nar

row

ing students’ average scores increased more than those

for White students.

Connecticut MissouriDistrict of Columbia New JerseyDelaware New YorkDoDEA OregonMassachusetts Rhode IslandMichigan

In Rhode Island, the gap widened between 2007 and 2009 as scores for White students increased, while

iden

ing

scores for Hispanic students did not change signifi-cantly.

W In Texas, the gap widened between 2007 and 2009 as scores for Hispanic students decreased, while scores for White students did not change significantly.

In Texas, scores for Hispanic students decreased, while scores for White students did not change significantly.

Figure 14 displays population percentages for Hispanic and White fourth-graders in each state for 2009 and the first year of the state’s participation in the assessment.

Hispanic-White mathematics gap data are not available for Maine, Mississippi, North Dakota, Vermont, or West Virginia.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

21

Figure 14. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by state: Various years, 1990–2009

Scale score

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

1990n1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1996n 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009 1992n 1996n 2003 2005 2007 2009

1992n 1996n 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2003 2005 2007 2009

219*

199*

227* 230* 233*243* 246* 248 248

201* 204* 207*221* 225* 227 227

218* 221* 227* 232* 235 238 237

218 220

232*242* 244* 247 249

228 227 232 232

225* 228* 230*241 243 246 243

203* 202* 204*217 218 220 220

217* 223* 225*237* 242 245 245

221*229 230 233

221* 223* 228*243* 245 247 247

190* 196* 201*216* 219 218 219

227* 232*243* 247* 249 252

204* 208*217* 223* 224 228

235* 240* 242*250* 250* 252 253

200* 201*210*

223 223 223 227

226* 225*244* 249 249 249

193*

226* 229 234 231

251* 248* 254* 262* 266 262*270

195* 196* 190*205*

215* 220* 227224* 227*

243* 247 250 250

208* 208*

232* 233* 238 238231* 233*

242* 245 246 245

214* 219*234 235 233 235

National average

National average

National average

National average

Alaska(2009: Hispanic 7%, White 50%)(1996: Hispanic 3%, White 66%)

Arizona(2009: Hispanic 45%, White 40%)(1992: Hispanic 23%, White 62%)

Arkansas(2009: Hispanic 8%, White 66%)(1992: Hispanic #, White 75%)

California(2009: Hispanic 51%, White 28%)(1992: Hispanic 30%, White 50%)

Colorado(2009: Hispanic 29%, White 61%)(1992: Hispanic 17%, White 73%)

Connecticut(2009: Hispanic 17%, White 66%)(1992: Hispanic 10%, White 76%)

Delaware(2009: Hispanic 12%, White 51%)(1992: Hispanic 2%, White 70%)

District of Columbia(2009: Hispanic 11%, White 7%)(1992: Hispanic 3%, White 5%)

DoDEA3

(2009: Hispanic 16%, White 49%)(1996: Hispanic 9%, White 49%)

Alabama(2009: Hispanic 4%, White 61%)(1992: Hispanic #, White 65%)

NATIONAL1

(2009: Hispanic 22%, White 54%)(1992: Hispanic 7%, White 72%)2

1517 1714

232622

2426

151316 1227 23 29 282527272731

252526 23

23 23292727 26

3239*34*2019 16 18

31*

16 1911 14 121342 43

57* 5164*5156*

20 1721 21 212126*2626*19

Florida(2009: Hispanic 25%, White 46%)(1992: Hispanic 12%, White 63%)

13 101071417*

See notes at end of figure.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


22

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Figure 14. Mathematics achievement score gaps between Hispanic and White public school students at grade 4, by state: Various years, 1990–2009—Continued

Scale score

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

1990n1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 2000 2003 2005 2007 2009 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

238*

235*244* 245* 248 249

211* 218* 219* 223 227

224*232* 235* 242* 245 249 247

226 230 233 230

231* 230* 233*241* 242* 245 245

222 222 230 223

237*246* 249 252 251

213*230 234 234 233 217* 222* 223*

231* 234* 238* 241

221 227

218* 221*230*

242 241 240 241

234 230

228* 234* 236* 244* 250* 251 255

207*216* 216*

227* 232 233 238232* 232* 230*

241* 243 245

222* 226* 227*238* 241* 244 247

206* 210*219 219 224 230

223* 227*238* 245 245

224

244

199*207*

217*226 225

228* 224* 230*241* 243 246 247

205*217* 219*

229 229 231219*

199*

227* 230* 233*243* 246* 248 248

201* 204* 207*221* 225* 227 227

NATIONAL1


21 21 212126*2626*19

National average

National average

National average

National average

19 21 19212024

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

Georgia(2009: Hispanic 11%, White 47%)(1992: Hispanic 1%, White 60%)

Hawaii(2009: Hispanic 3%, White 14%)(1992: Hispanic 2%, White 23%)

Idaho(2009: Hispanic 14%, White 81%)(1992: Hispanic 6%, White 92%)

Illinois(2009: Hispanic 22%, White 51%)(2000: Hispanic 20%, White 56%)

Indiana(2009: Hispanic 7%, White 76%)(1992: Hispanic 2%, White 87%)

Iowa(2009: Hispanic 8%, White 84%)(1992: Hispanic 1%, White 95%)

Kansas(2009: Hispanic 15%, White 69%)(2000: Hispanic 9%, White 79%)

Kentucky(2009: Hispanic 3%, White 83%)(1992: Hispanic #, White 90%)

Louisiana(2009: Hispanic 4%, White 47%)(1992: Hispanic 1%, White 53%)

Maine(2009: Hispanic 1%, White 94%)(1992: Hispanic #, White 98%)

1714 15221319

20 172316 16

18

25 212624 27161516 16

152018

181816

2416 16 14

6 10

22

Maryland(2009: Hispanic 11%, White 48%)(1992: Hispanic 2%, White 62%)

17 1818 17201721

Gap data not available


Stat

e n G

rade

4

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

23


0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009

1992n 1996n 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009 1992n 2003 2005 2007 2009

231* 232* 239* 247* 252* 257 258

197*206* 203*

222* 225* 231 232

232*227*237* 244 245 244 243

205*223 224 230 227

231* 235* 238*246* 251* 252 255

220* 223* 229 232

219* 221* 222*236* 238 239 241

227* 230* 233* 240* 240* 245 245

220* 221*234 237

231* 231* 238* 243* 247 247

236 234 241 241

228* 231* 230*241* 244 244 245

203* 198* 205*213* 219* 220 224

224* 226*236* 240* 243 245

204* 207*216* 219* 221 227

230*244* 246* 250 252

225* 226 232 234

236* 239*248* 251* 255 255

204* 206*224* 230 234 232

224* 227* 227*237* 238* 242 245

203* 204* 207*217* 218* 222 224

233* 238*246 247 251* 248

201*

228*

197*207*

221* 226* 230 231

National average

National average

National average

National average

14 16222128*

23282623

11 8 669220*19*

26 24 2133

29*25252120 21

21202021

212023

2019 182019 17

21

21 242133*

2332*

2121 1725*30*33*32*

26 262736

252734*

New York(2009: Hispanic 20%, White 52%)(1992: Hispanic 17%, White 63%)

Massachusetts(2009: Hispanic 17%, White 68%)(1992: Hispanic 4%, White 83%)

Michigan(2009: Hispanic 5%, White 71%)(1992: Hispanic 3%, White 79%)

Minnesota(2009: Hispanic 7%, White 76%)(1992: Hispanic 2%, White 91%)

Mississippi(2009: Hispanic 2%, White 45%)(1992: Hispanic #, White 42%)

Missouri(2009: Hispanic 4%, White 76%)(1992: Hispanic 1%, White 83%)

Montana(2009: Hispanic 3%, White 83%)(1996: Hispanic 2%, White 85%)

Nebraska(2009: Hispanic 16%, White 73%)(1992: Hispanic 3%, White 90%)

Nevada(2009: Hispanic 39%, White 42%)(1996: Hispanic 16%, White 66%)

New Hampshire(2009: Hispanic 4%, White 91%)(1992: Hispanic 1%, White 96%)

New Jersey(2009: Hispanic 21%, White 55%)(1992: Hispanic 11%, White 69%)

New Mexico(2009: Hispanic 58%, White 28%)(1992: Hispanic 45%, White 45%)

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap



1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


24

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II



Scale score

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

1990n1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 2000 2003 2005 2007 2009 1992n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009

2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

219*

199*

227* 230* 233*243* 246* 248 248

201* 204* 207*221* 225* 227 227

NATIONAL1


21 21 212126*2626*19

223*233* 238*

251 250* 251 254

220*235 234 235 236

230* 232* 232*240* 245* 248 248

222*235* 243* 248 250 249

225 231 231 233224* 229* 235* 240 242 241

207* 211*220* 226 227 229

226* 227*240* 243 241 243

197* 202*218 218 217 221

230* 231*243* 247 249 249

201* 202*216* 220

229 227221* 225* 232* 239* 241* 242* 247

186* 191* 197*207* 211*

220 219

225* 224*233*

246 250* 248 245

232 236227 232

241* 245 245 247

223* 228 233 217*226* 227* 235* 238 240 239

218229 222 225

230*240* 241* 248* 254 253 254

208*216* 223* 230* 235 236* 233

National average

National average

National average

National average

16 18161718

151318 15

1724 2112 25

29*22

25

27 27 19 222929

22*293235*34

28

35*

18 18 13 1417 9 18

20 131413

191818 17*20

2422

North Carolina(2009: Hispanic 11%, White 54%)(1992: Hispanic 1%, White 65%)

North Dakota(2009: Hispanic 2%, White 86%)(1992: Hispanic 1%, White 95%)

Ohio(2009: Hispanic 3%, White 72%)(1992: Hispanic 1%, White 86%)

Oklahoma(2009: Hispanic 10%, White 58%)(1992: Hispanic 3%, White 77%)

Oregon(2009: Hispanic 17%, White 69%)(1996: Hispanic 6%, White 85%)

Pennsylvania(2009: Hispanic 9%, White 71%)(1992: Hispanic 3%, White 81%)

Rhode Island(2009: Hispanic 18%, White 68%)(1992: Hispanic 7%, White 82%)

South Carolina(2009: Hispanic 6%, White 55%)(1992: Hispanic #, White 58%)

South Dakota(2009: Hispanic 3%, White 80%)(2003: Hispanic 2%, White 84%)

Tennessee(2009: Hispanic 5%, White 69%)(1992: Hispanic #, White 73%)

Texas(2009: Hispanic 51%, White 31%)(1992: Hispanic 34%, White 49%)

18 18 18 16


Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

Stat

e n G

rade

4


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

25


Scale score

0

175

225

275

500

0

175

225

275

500

0

175

225

275

500

1992n 1996n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009

1996n 2003 2005 2007 2009 1992n 1996n 2000 2003 2005 2007 2009 1992n 1996n 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009

225* 228* 230* 238* 242* 244 246

206* 204* 205*216 220 220 219

228* 230* 237*246* 247* 251 251

214224* 230 230 235 234

229*242* 246 248 247

204*223 224 225 227

225* 232*242* 244* 247 248

216*224* 224* 231* 231* 237* 233 233* 236* 243* 247* 250 250

208* 211*221 224 229 228

227* 225* 231*243 245 246* 244

216*207* 214*

229 234 229 231

National average

National average

National average

1517 17161315

2123222522

25

Virginia(2009: Hispanic 8%, White 56%)(1992: Hispanic 2%, White 71%)

23 20

27242222252420

13171114171811

221925

Utah(2009: Hispanic 16%, White 77%)(1992: Hispanic 4%, White 93%)

Vermont(2009: Hispanic 1%, White 94%)(1996: Hispanic #, White 97%)

Washington(2009: Hispanic 18%, White 62%)(1996: Hispanic 6%, White 79%)

West Virginia(2009: Hispanic 1%, White 92%)(1992: Hispanic #, White 96%)

Wisconsin(2009: Hispanic 9%, White 75%)(1992: Hispanic 2%, White 87%)

Wyoming(2009: Hispanic 11%, White 84%)(1992: Hispanic 6%, White 90%)



Hispanic

White

Gap

Hispanic

White

Gap

Hispanic

White

Gap

n Accommodations were not permitted for this assessment.# Rounds to zero.

* Significantly different (p<.05) from 2009.1 National results for assessments prior to 2002 are based on the national sample, not on aggregated state samples.2 Hispanic and White percentages are based on students tested in the first assessment year for the state and in 2009.3 Department of Defense Education Activity (overseas and domestic schools). Before 2005, DoDEA overseas and domestic schools were separate jurisdictions in NAEP. Pre-2005 data presented here were recalculated for comparability.NOTE: Score gaps are calculated based on differences between unrounded average scores. Where data are not present, the jurisdiction did not participate or did not meet the minimum participa-tion guidelines for reporting. Comparative performance results may be affected by changes in exclusion rates for students with disabilities and English language learners in the NAEP samples.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1990–2009 Mathematics Assessments.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


26

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Stat

e n G

rade

8

State and national mathematics achievement gaps at grade 8, 2009Fifteen states had a smaller gap than the nation’s 26-point students was not statistically significant, and thus there was Hispanic-White gap in 2009 (Alaska, Arkansas, Delaware, no measurable Hispanic-White gap for grade 8 mathemat-DoDEA, Florida, Georgia, Hawaii, Indiana, Kentucky, ics in those states in 2009. Among the remaining states, the Michigan, Missouri, Oklahoma, Tennessee, Virginia, and Hispanic-White gap ranged from 10 points in Kentucky to Wyoming) and six had a gap that was larger (California, 34 points in Connecticut and Massachusetts.Colorado, Connecticut, New York, Rhode Island, and

Among the 15 states with a gap smaller than the national Washington). In 24 states, the gap was not significantly dif-

gap of 26 points, in one state, Delaware, scores for both ferent from the nation’s gap (figure 15). Asterisks indicate

Hispanic and White students were higher than the nation-a significant difference from the national average score or

al averages for those students. In eight of these states— the national average gap.

Alaska, DoDEA, Florida, Georgia, Hawaii, Indiana The eighth-grade mathematics gap in 2009 was statisti- Missouri, and Virginia—the average score for Hispanic cally significant in 43 of the 45 states for which data could students was above the national average for those students.be reported. In Hawaii and Missouri, the 6-point differ-

Among the six states where the gap was larger than ence between the average scores for Hispanic and White

the national gap, scores for Hispanic students were not

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

27

measurably different from the national average, while scores for White students were higher in three states—Colorado, Connecticut, and Washington. In New York, scores for Hispanic students were not measurably different from the national average for Hispanic students and scores for White students were not measurably different from the national average for White students. In California, scores for Hispanic students were below the national average for Hispanic students and scores for White students were not measurably different from the national average for White students. In Rhode Island, scores for both groups of students were below the national average.

Figure 15. The Hispanic-White achievement score gap in mathematics for public school students at grade 8, by state: 2009

Jurisdictions

5000

Scale score

240 250 260 270 280 290 300 310 320

WhiteHispanic Gap

WyomingWisconsin

WashingtonVirginia

UtahTexas



OregonOklahoma

OhioNorth Carolina


New HampshireNevada


MinnesotaMichigan

MassachusettsMarylandKentucky

KansasIowa



DelawareConnecticut

ColoradoCaliforniaArkansas

ArizonaAlaska

Alabama

National Public 266

260275*

265269

256*267

263278*

281*274*

270*276*

264269

273*266

274*272

275*271

269269

284*278*

262262*

270272

262*262

274*267

263264

266255*

269268

270277*

259*274*

264268269

292

280* 293

292 284*

289 299*

298* 294* 294

289* 289

282* 292

294 291

287* 294

282* 303*

305* 286*

300* 290*

296* 291

287* 293

302* 288*

294 297*

291 282*

290 294

286* 293

295* 282*

301* 289*

294 295*

294* 289*

20 18* 26 15* 33* 32* 34* 16* 13* 15* 19* 6* 28 25 18* 21 20 10* 28 34 17* 31 6* 17 29 25 23 30 26 32* 23 24 19* 26 28 31* 23 27 12* 24 30 19* 32* 26 20*

26

* Significantly different (p<.05) from the nation (public) when comparing one state to the nation at a time.1 Department of Defense Education Activity (overseas and domestic schools).NOTE: Reporting standards not met for District of Columbia, Louisiana, Maine, Mississippi, North Dakota, Vermont and West Virginia because their Hispanic or White student population size was insufficient for comparison. They are not included in the figure. Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Mathematics Assessment.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


28

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Trends in state mathematics achievement gaps at grade 8The national Hispanic-White achievement gap was not significantly narrower when comparing 2009 to 1990 despite higher average scores for both Hispanic and White students in 2009 (figure 16, national results). Average scores were also higher for both Hispanic and White students in 2009 than in 2007, although again the gap did not narrow.

In 14 of the 15 states for which 1990 data were avail-able, mathematics scores of both Hispanic and White eighth-graders were higher in 2009 than in 1990. In Connecticut, Delaware, Hawaii, Missouri, and Rhode Island, the Hispanic-White gap was narrower in 2009 than in the first assessment year for which reportable results for both groups were available: Connecticut (1990); Delaware (2003); Hawaii (1996); Missouri (2007); Rhode Island (1990).

Maryland’s gap was wider in 2009 than in 1990, as White students’ scores increased more than those of their Hispanic peers. In Utah, the gap was wider in 2009 than in 1992, the state’s first assessment year, as scores for White students increased while scores for Hispanic students did not change significantly.

Between 2007 and 2009, gaps narrowed in Arkansas and Delaware as scores for Hispanic eighth-graders increased while scores of White students did not change significantly.

Stat

e n G

rade

8

Changing of the Gap

In Connecticut, Delaware, Hawaii, Missouri, and Rhode Island, the gap was narrower in 2009 than in

win

g the first assessment year for which reliable results for both groups were available as scores for Hispanic stu-

Nar

ro dents increased more than scores for White students.

In Arkansas and Delaware the gap narrowed between 2007 and 2009 as Hispanic students’ average scores increased while those for White students did not.

ng In Maryland and Utah the gap was wider in 2009 than

ni in the first assessment year for each state as White id

estudents’ scores increased more than those of their

W Hispanic peers.

In Connecticut and Nevada, scores increased for both Hispanic and White eighth-grade students between 2007 and 2009, though the gaps did not narrow significantly.

Figure 16 displays population percentages for Hispanic and White eighth-graders in each state for 2009 and the first year of the state’s participation in the assessment.

Hispanic-White mathematics gap data are not avail-able for Louisiana, Maine, Mississippi, North Dakota, Vermont, or West Virginia.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

29

Figure 16. Mathematics achievement score gaps between Hispanic and White public school students at grade 8, by state: Various years, 1990–2009

Scale score

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009 1996n 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 1996n 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2003 2005 2007 2009

269* 276* 280* 283* 287* 288* 290* 292

245* 247* 250* 252* 258* 261* 264* 266262* 264* 270* 275* 274* 276* 278 280

249260

285* 290* 288* 294 293

263*272 274 275

270* 274* 277* 281* 284* 288 289 292

241* 247* 248* 248*258* 260* 262 265

264* 264* 269* 268* 275* 281 282 284

248*266

256*269

270* 275* 277* 277* 283* 284* 287 289

236* 239* 245* 242*250* 254 256 256

273* 278* 282*292* 292* 296 299

247* 252* 255* 259* 260* 264 267

277* 283* 287* 291* 293* 293* 293* 298

235* 239*251* 249*

259 254* 254*263

268* 272* 275*287* 291* 294 294

257*268* 267*

278

300317

250*

226*236*

246* 252* 251*265

282* 286* 293 292 291 294

267* 267*278 280 282 281 265* 272* 277*

286 286 289 289

246* 246*254*

264* 265* 270 274

National average

National average

National average

National average

NATIONAL1


29 20


2017 1827


263329282928 27 26


27 25* 1515

Arkansas(2009: Hispanic 8%, White 69%)(1990: Hispanic 1%, White 75%)

34 36 32 3531 332933


262732 32313327


42* 44* 36 42 393934 34


1627*30*23*



6564

DoDEA3


24 3029 31* 26 262628*


19 26* 23*21*22 18 1516 19 15 12 10 13

White

White

GapHispanic

WhiteGapHispanic

WhiteGapHispanic

GapHispanic


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


30

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II



Scale score

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 2000 2003 2005 2007 2009 1990n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 20091990n1992n 1996n 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 20091990n1992n 1996n 2000 2003 2005 2007 2009

270* 270* 276* 279* 284* 284* 288 289

262* 258* 266 270

273* 277* 280* 284* 284* 287* 292

250* 255* 250* 251*261 264 264

270*285* 289* 289* 291 294

238*258* 259* 265 265 269

270* 273* 280* 285* 286* 286* 290 291

261* 261* 267 273

279* 284 285 287 286 288 287

255264 261 266

287* 290* 289* 295 294

263* 263* 266* 269 274259* 264* 269* 272* 277* 276* 282 282

272

259* 263* 266*275* 281 281 283 283 279* 284 281* 282* 281* 287 287

272* 278* 284* 286* 289* 292*300 303

254*263 262* 262*

272 275

259* 263*276* 274* 273* 277* 278 282

253*263 257* 264

276

National average

National average

National average

National average

NATIONAL1


24 3029 31* 26 262628*

23 22 3024 282333

21 22 1927

3327 30 26 2524 23 182526

15 62023* 9

21282232 24 26 202427 10

18*23 27 30

28 28

White

White

GapHispanic

WhiteGapHispanic

WhiteGapHispanic

GapHispanic









Gap data not available Gap data not available




269* 276* 280* 283* 287* 288* 290* 292

245* 247* 250* 252* 258* 261* 264* 266


Stat

e n G

rade

8

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

31


Scale score

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

2000 2003 2005 2007 2009 1990n1992n 2003 2005 2007 2009

277* 283* 284* 292* 297*305 305

239* 239* 246*255*

265 270 271

270* 276*284 285 286 285 285 286

252267 265 259

269

277* 284* 287* 290* 295* 296* 297 300

262 263 269 269

262* 265* 268* 275* 279 279 279 275* 278* 277* 284* 284* 288 290

270*284

282* 286* 288* 289* 290* 291* 296

278

279* 281* 285* 285* 287* 289 291 291

256 260242*

255 261 261 262

273* 278* 280* 282* 287

249* 250* 256* 257* 262

279* 283*292* 295* 298 302

242* 245*262* 264* 271 272

271* 272* 277* 274*282* 279* 285 288

247* 248* 252* 251* 254* 255* 260 262

273* 278*287* 286* 289* 293

264 270

273* 280* 283* 284*293 290 290 294

238* 241* 244* 251*262 262 264 262

National average

National average

National average

National average

1925 2620 17 28 313433

17* 6

25 33 29 29284326

24 29 252524

27 3030303737 2428 25 26

22252424

4438 3735 343237

35 39 39283134 26

23

32

17

24



Minnesota(2009: Hispanic 5%, White 79%)(1990: Hispanic #, White 93%)











White

White

GapHispanic

WhiteGapHispanic

WhiteGapHispanic

GapHispanic


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


32

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II



Scale score

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

1990n1992n 2000 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 2003 2005 2007 2009

2003 2005 2007 2009

1990n1992n 2000 2003 2005 2007 2009 1990n 1996n 2000 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 2009

1992n 1996n 2000 2003 2005 2007 20091990n1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 20091992n 1996n 2000 2003 2005 2007 2009

261* 266*277*

287* 294 292* 295 297

263* 265*273 274

284* 284* 286* 285* 290* 290* 295 296

268* 274*285* 287* 289 291 291

270259

276267

268* 272* 274* 278* 278* 280 282

260 258 257 259 263

273* 278* 284* 284* 287* 289 290

256 257248*

258 257* 261 264

272* 276*285* 287* 293 294

253*267 264 266

273* 273* 277*291 294 293 293

269 272 269

288* 291* 292* 295

269 268

266* 270* 269* 277* 278* 282 282

264 270

272* 278* 284* 286* 290* 295* 300 301

245* 249* 255* 262* 267* 271* 277 277

265* 271* 275* 275* 280* 281* 284 286

227* 227*238* 240* 245* 244* 251 255

National average

National average

National average

National average

NATIONAL1


24 3029 31* 26 262628*2230 2326

17 30 15 2414 21 21 1921 28 263026362116

32 29 28

2723

20

38* 43* 36 3536 33 3137*21 2325

18 1228 29 29

24232423 24










Tennessee(2009: Hispanic 3%, White 70%)(1992: Hispanic #, White 77%)



White

White

GapHispanic

WhiteGapHispanic

WhiteGapHispanic

GapHispanic

269* 276* 280* 283* 287* 288* 290* 292

245* 247* 250* 252* 258* 261* 264* 266


Stat

e n G

rade

8

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Ma

th

em

at

ics

33


Scale score

0

225

275

325

500

0

225

275

325

500

0

225

275

325

500

1992n 1996n 2000 2003 2005 2007 2009 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2000 2003 2005 2007 2009

1996n 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 2009

1990n1992n 1996n 2000 2003 2005 2007 2009 1990n1992n 1996n 2003 2005 2007 2009

276* 278* 277*285* 283* 286* 289

253 257244* 249* 255 256 259

280* 281* 286* 288* 292 293

271* 275* 279* 283* 290 293 296 294

263* 268 270 275 274

281* 285* 289* 291* 295

248*263 262 263 264

256* 260* 265* 267* 271 270 271 271279* 282* 288* 290* 291 292 294

262 265 268 268

274* 277* 277* 278*286* 284* 290 289

257* 262 256* 257*265 265

274 269

National average

National average

National average


2122* 3331 302836

16 2015 2116 202020

21 192320 22

24 2628 26



White

White

GapHispanic

GapHispanic

332722* 27 32

Vermont(2009: Hispanic 1%, White 94%)(1996: Hispanic 1%, White 96%)






WhiteGapHispanic


* Significantly different (p<.05) from 2009.1 National results for assessments prior to 2002 are based on the national sample, not on aggregated state samples.2 Hispanic and White percentages are based on students tested in the first assessment year for the state and in 2009.3 Department of Defense Education Activity (overseas and domestic schools). Before 2005, DoDEA overseas and domestic schools were separate jurisdictions in NAEP. Pre-2005 data presented here were recalculated for comparability.NOTE: Score gaps are calculated based on differences between unrounded average scores. Where data are not present, the jurisdiction did not participate or did not meet the minimum participa-tion guidelines for reporting. Comparative performance results may be affected by changes in exclusion rates for students with disabilities and English language learners in the NAEP samples.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1990–2009 Mathematics Assessments.

34

Mathematics Summary

National trends in mathematics, 1990–2009n The mathematics achievement gap between Hispanic

and White students in 2009 was 21 points at grade 4 and 26 points at grade 8. There was no significant difference when comparing these gaps to the comparable gaps in 2007 or 1990 (figures 5 and 6).

n Neither males nor females demonstrated a significant change in the Hispanic-White gap when comparing 2009 to 2007 or to 1990 (figures 7 and 8).

n Among eighth-grade students who were eligible for free or reduced-price school lunch through the National School Lunch Program, the Hispanic-White achieve-ment gap was narrower in 2009 than in 2003. The corre-sponding gap for fourth-grade students did not change significantly over the same period (figure 9 and 10).

n The gap between White and non-ELL Hispanic stu-dents at both grades was narrower in 2009 than in 1996, the first year for which data was available (figures 11 and 12).

State achievement gaps in mathematics, compared to the nation, 2009n At grade 4, eleven states had a smaller Hispanic-White

gap than the nation, and six had a gap that was larger (figure 13).

n At grade 8, fifteen states had a smaller Hispanic-White gap than the nation, and six a gap that was larger (figure 15).

State trends in mathematics, 1990–2009n At grade 4, of the 21 states for which 1992 data was

available, six states had a narrower Hispanic-White achievement gap in 2009 than in 1992. Two states had a wider Hispanic-White achievement gap in 2009 than in 2007 (figure 14).

n At grade 8, of the 15 states for which 1990 data was avail-able, two states had a narrower gap in 2009 than in 1990, and one state had a gap that was wider. Additionally, three states had a narrower gap in 2009 than in 2007 (figure 16).

Ma

th

em

at

ics

35

36

National Results for Hispanic and White Fourth- and Eighth-Graders

Reading scores and achievement gaps in the nation, 1992–2009Average fourth-grade reading scores for the nation were higher in 2009 than in 1992 for both Hispanic and White public school students (figure 17). The 25-point gap in 2009 was not significantly different from the 28-point gap in 1992. From 2007 to 2009, scores did not change signifi-cantly for either group. The 26-point gap in 2007 was not significantly different from the 25-point gap in 2009.

Average reading scores for both Hispanic and White eighth-graders were higher in 2009 than in 1992 and in 2007 (figure 18). The 24-point gap in 2009 was not signifi-cantly different from either the 27-point gap in 1992 or the 25-point gap in 2007.

Re

ad

ing

37

Figure 17. Reading achievement score gaps between Hispanic and White public school students at grade 4: Various years, 1992–2009

1992n 1994n 1998 2000 2002 2003 2005 2007 2009

Scale score

0

160

210

260

500

229230228*227*227*223*223*222*223*

204204201*199*199*188*192*186*

194*

GapWhite

Hispanic28 36* 31 35* 28*28 25 2626

National average


* Significantly different (p<.05) from 2009.NOTE: Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1992–2009 Reading Assessments.

Figure 18. Reading achievement score gaps between Hispanic and White public school students at grade 8: Various years, 1992–2009

1992n 1994n 1998 2002 2003 2005 2007 2009

Scale score

0

200

250

300

500

265* 265* 268* 271 270* 269* 270* 271

238* 239* 241* 245 244* 245* 246* 248

GapWhite

Hispanic27 25 27 27*26 24 2524

National average


* Significantly different (p<.05) from 2009.NOTE: Data were not collected at grade 8 in 2000. Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1992–2009 Reading Assessments.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 199

II

II

II I I I I I

9

Nati

onal

n G

rade

4 &

8

9

38

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 19 9 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Reading scores and achievement gaps by gender, 1992-2009

Average reading scores were higher in 2009 than in 1992 From 2007 to 2009, scores did not change for any of these for the nation’s Hispanic and White students at both the student groups at grade 4. At grade 8, scores increased fourth and eighth grades, regardless of gender (figures 19 for Hispanic and White students, regardless of gender. and 20). The Hispanic-White reading gap did not change There were no significant changes in the size of the gaps significantly for either male or female students at either at either grade.grade when comparing 2009 to 1992.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Re

ad

ing

39

Figure 19. Reading achievement score gaps between Hispanic and White public school students at grade 4, by gender: Various years, 1992–2009

Scale score

0

160

210

260

500

1992n 1994n 1998 2000 2002 2003 2005 2007 2009 1992n 1994n 1998 2000 2002 2003 2005 2007 2009

232233231*231*231*228*227*227*227*226227225*223*224*218*219*217*219*

207207204*203*202*194*193*191*

199*201201199*196*197*182*

191*181*

190*

WhiteGapHispanic

26 252627*2736*2936*2927 2628*34* 34* 28 25

36*28

Male Female



Figure 20. Reading achievement score gaps between Hispanic and White public school students at grade 8, by gender: Various years, 1992–2009

Scale score

0

200

250

300

500

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

276275*275*276276276272*272*267265*264*265*266261*257*258*

252250*249*248*249248245*243*243241*241*240*241235*234*234*

WhiteGapHispanic

24 2323252526232426 2528*27 27 242729

Male Female

n Accommodations were not permitted for this assessment.* Significantly different (p<.05) from 2009.NOTE: Data were not collected at grade 8 in 2000. Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1992–2009 Reading Assessments.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


40

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Reading scores and gaps by family income, 2003–2009 The National Assessment of Educational Progress (NAEP) uses student eligibility for free or reduced-price school lunch as an indicator of family income. Eligible students come from families with an income no more than 185 percent of the federal poverty level. As table 3 indicates, in 2009, 76 percent of fourth-grade and 72 per-cent of eighth-grade Hispanic public school students were eligible for free or reduced-price school lunch. For White public school students, 29 percent of fourth-graders and 24 percent of eighth-graders were eligible. Trends in school lunch eligibility for the NAEP reading assessment are reported from 2003.

At grade 4, reading scores were higher in 2009 than in 2003 for both not eligible and eligible Hispanic and White pub-lic school students (figure 21). Additionally, the Hispanic-White gap did not change significantly for not eligible students, but did narrow for eligible students, falling from 17 points to 15 points. From 2007 to 2009, there were no significant changes in scores or gaps for White or Hispanic students at grade 4, regardless of eligibility.

At grade 8, scores in 2009 were higher than in 2003 for White students not eligible for free or reduced-price school

Nati

onal

n G

rade

4 &

8

Table 3. Percentage of public school students assessed in NAEP reading eligible for the National School Lunch Program, by race/ethnicity and grade: Various years, 2003–2009

Hispanic WhiteGrade 4 2009 76 29 2007 74 26* 2005 76 26* 2003 74 25*Grade 8 2009 72 24 2007 70 24 2005 68* 23* 2003 67* 20*

* Significantly different (p<.05) from 2009.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 2003–2009 Reading Assessments.

lunch and for eligible Hispanic students (figure 22). The not eligible gap did not change significantly, but the eligi-ble gap narrowed from 18 points to 14 points since eligible Hispanic students’ scores increased, while eligible White students’ scores did not.

From 2007 to 2009, scores increased for not eligible White eighth-graders; the gap did not change significantly. Scores for eligible Hispanic students also increased, while scores for eligible White students did not, and the gap narrowed from 16 points to 14 points (figure 22).

In 2009, the reading performance gap for Hispanic eligible and not eligible students was smaller than the gap for White eligible and not eligible students, at both grades. At grade 4, White not eligible students had an average score of 235 (figure 21 left side of graph), while White eligible students had a score of 215 (figure 21 right side of graph), resulting

NAEP collects data on students’ eligibility for the National School Lunch Program (NSLP)—sometimes referred to as the free or reduced-price school lunch program—as an indicator of family economic status. Eligibility for free or reduced-price lunches is based on students’ family income in relation to the federally established poverty level.

Not eligible: Students who are not eligible for the program because their family’s income is above 185 percent of the poverty level.

Eligible: Students who are eligible for either reduced-price lunch because their family’s income is between 130 per-cent and 185 percent of the poverty level, or for free lunch, because their family’s income is below 130 percent of the poverty level.

As a result of improvements in the quality of the data on students’ eligibility for NSLP, the percentage of students for whom information was not available has decreased in comparison to the percentages reported prior to the 2003 assessment. Therefore, trend comparisons are only made back to 2003 in this report.

Eligibility for free or reduced-price lunch

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Re

ad

ing

41

in a gap of 20 points. Hispanic not eligible students had an average score of 217, while Hispanic eligible students had a score of 200, resulting in a gap of 16 points (using unrounded numbers), smaller than the gap between White students.

At grade 8, White not eligible students had an average score of 276 (figure 21 left side of graph), while White

eligible students had a score of 258 (figure 22 right side of graph), resulting in a gap of 18 points. Hispanic not eli-gible students had an average score of 259, while Hispanic eligible students had a score of 244, resulting in a gap of 15 points, which was also smaller than the gap for White students.

Figure 21. Reading achievement score gaps between Hispanic and White public school students at grade 4, by eligibility for free or reduced-price school lunch: Various years, 2003–2009

Scale score

0

160

210

260

500

2003 2005 2007 2009 2003 2005 2007 2009

215215214212*

235235233*232*

200199197*195*

217217217213*

WhiteGapHispanic


16*17* 17* 16 15

18 1820


Figure 22. Reading achievement score gaps between Hispanic and White public school students at grade 8, by eligibility for free or reduced-price school lunch: Various years, 2003–2009

Scale score

0

200

250

300

500

2003 2005 2007 2009 2003 2005 2007 2009

258258257257276274*273*274*

244242*241*239*

259256255*256

WhiteGapHispanic


1818* 16 16* 14

18 1718


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


42

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Trends in reading scores and achievement gaps by ELL status at grades 4 and 8, 1998-2009The 15-point gap between grade 4 White students (both ELL and non-ELL White students) and non-ELL Hispanic students in 2009 was narrower than the 24-point gap in 1998, the first year for which data are available (figure 23). Scores rose for both groups, but the increase for non-ELL Hispanic students was larger. From 2007 to 2009, there was no significant change in the scores of either group or in the gap. (For percentages of ELL students in fourth and eighth grade, see table 4.)

Table 4. Percentage of public school English language learner students assessed in NAEP reading, by race/ethnicity and grade: Various years, 1998–2009

Hispanic WhiteGrade 4 2009 2007 2005 2003 2002 2000 1998

35 37 37 37 34 33 22

11111#1

Grade 8 2009 2007 2005 2003 2002 1998

20 24 24 24 22 15

#1111#

# Rounds to zero.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1998–2009 Reading Assessments.

Nati

onal

n G

rade

4 &

8

Within the Hispanic student population, the 29-point gap between ELL and non-ELL Hispanic students in 2009 was not significantly different from either 1998 or 2007. Scores for both groups were higher in 2009 than in 1998 but not significantly different from 2007.

At grade 4 in 2009, the 29-point gap between non-ELL Hispanic students and ELL Hispanic students was larger than both the 15-point gap between White and non-ELL Hispanic students (figure 23) and the 25-point gap between all Hispanic and White students (figure 17).

At grade 8, average reading scores for White students and non-ELL Hispanic students were higher in 2009 than in 1998, the first year for which data are available (figure 24). They were also higher in 2009 than in 2007. The 15-point gap in 2009 was narrower than the 22-point gap in 1998, but was not measurably different from the 16-point gap in 2007.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Reading

43

Within the Hispanic eighth-grade student population, the 39-point gap between ELL and non-ELL Hispanic stu-dents in 2009 was wider than the gap in either 1998 or 2007. The average score for non-ELL Hispanic students in 2009

was higher than in either comparison year, while the 2009 score for ELL Hispanic students was not significantly dif-ferent from their scores in either 1998 or 2007 (figure 24).

Figure 23. Reading achievement score gaps between Hispanic and White public school students at grade 4, by English language learner status: Various years, 1998–2009

1998 2002 2003 2005 2007 2009 1998 20022000 2000 2003 2005 2007 2009

Scale score

0

160

210

260

500

229230228*227*227*223*223*

214215211*209*210*199*199*

214215211*209*210*199*199*

185184184183179166*166*


24* 23* 18* 16*18* 14 15GapELL Hispanic

Non-ELL Hispanic

333034

25* 27 31 29


* Significantly different (p<.05) from 2009.NOTE: Score gaps are calculated based on differences between unrounded average scores. White includes ELL and non-ELL White students.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1998–2009 Reading Assessments.

Figure 24. Reading achievement score gaps between Hispanic and White public school students at grade 8, by English language learner status: Various years, 1998–2009

1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

Scale score

0

200

250

300

500

268* 271 270* 269* 270* 271

246* 252* 252* 253* 254* 256246* 252* 252* 253* 254* 256

215 222* 218 220* 219 217


22* 19* 17*19* 16 15GapELL Hispanic

Non-ELL Hispanic

32* 31* 34* 32* 35* 39


* Significantly different (p<.05) from 2009.NOTE: Data were not collected at grade 8 in 2000. Score gaps are calculated based on differences between unrounded average scores. White includes ELL and non-ELL White students.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1998–2009 Reading Assessments.

44

State Results for Hispanic and White Fourth- and Eighth-Graders

State and national reading achievement gaps at grade 4, 2009The NAEP state reading assessments were administered larger. In 29 states, the gap was not significantly different to public school students in fourth grade only in 1992 from the nation’s gap (figure 25). Asterisks indicate a sig-and 1994, and in fourth and eighth grade in 1998, 2002, nificant difference from the national average score or the 2003, 2005, 2007, and 2009. Before 2003, states were not national average gap.required to participate in NAEP in order to qualify for

In two states, Mississippi and Ohio, the difference in scores Title I education funds. Typically, 40 or more states par-

between Hispanic and White students was not statistically ticipated in each assessment prior to 2003. In 2003, 2005,

significant, and thus there was no measurable gap in those 2007, and 2009, all 50 states, the District of Columbia, and

two states. The difference was 12 points in Mississippi and the DoDEA schools participated.

15 points in Ohio. The fourth-grade reading gap in 2009 State results are presented in two ways. Comparisons of was statistically significant in the other 46 states for which fourth-grade reading gaps in 2009 between each state and data could be reported.the nation are presented in figure 25.

In 4 of the 13 states where the gap was smaller than Comparisons of the reading gaps within a state over time the national average—Delaware, DoDEA, Florida, and are presented in a series of small graphs in figure 26. At Maryland—scores for both Hispanic and White stu-the top left of each two-page spread, the reading scores dents were above the national average for those stu-and gaps for the nation are presented for reference. Each dents. In four states—Hawaii, Kentucky, Missouri, and state figure, as well as the national figure, also contains a Montana—scores for Hispanic students were higher than dashed gray line representing the national average for pub- they were nationally while scores for White students were lic school students. The data for the national averages are comparable. In three states—Alaska, South Dakota, and located in appendix B in table B-4. Wyoming—scores for Hispanic students were higher

than the national score while scores for White students Thirteen states had a smaller Hispanic-White gap than the

were lower.nation’s 25-point gap in 2009 and six had a gap that was

Re

ad

ing

45

Among states where the gap was larger than the national average, in Utah scores for both Hispanic and White students were below the national averages for those stu-dents. In California and Minnesota, the scores for Hispanic students were below the national average while the scores for White students were not measurably dif-ferent from it. In Colorado, Connecticut, and the District of Columbia, the score for Hispanic students was not measurably dif-ferent from the national average while the score for White students was higher than the national average.

In Mississippi the score for Hispanic stu-dents was not measurably different from the national average while the score for White students was below it. In Ohio the scores for both Hispanic and White stu-dents were not measurably different from the national averages for those students.

Figure 25. The Hispanic-White achievement score gap in reading for public school students at grade 4, by state: 2009

5000

Scale score

180 190 200 210 220 230 240 250 260

WyomingWisconsin

WashingtonVirginia

UtahTexas



OregonOklahoma

OhioNorth Carolina


New HampshireNevada


MississippiMinnesotaMichigan

MassachusettsMarylandLouisianaKentucky

KansasIowa


HawaiiGeorgiaFloridaDoDEA

District of ColumbiaDelaware

ConnecticutColorado

CaliforniaArkansas

ArizonaAlaska

Alabama

National Public 204

200215*

198*202

196*204205

216*207

223*223*

208215*

201203203

207210

215*206

221*211*

206194*

212216*

219*207

199*217*

213*201

210*204

215207

196*199200

205216*

202210*

194*214*

201202

212*

229

225* 226*

225* 224*

227 236*

238* 235*

256* 234*

233* 229

226 225*

231 227*

224* 229 228

219* 237*

241* 225*

230 225*

228 228 228

222* 230

237* 224*

233* 230 230

223* 223*

230 231

226 227*

224* 232

225* 234*

229 227*

224*

Jurisdictions WhiteHispanic

25 11* 27 22 31* 32* 33* 18* 49* 11* 10* 21 12* 24 28 24 16* 19 13* 13* 15* 30 19 36* 12 12* 9* 21 23 13 24 22 22 26 15 16

31 31 22 11* 22 22

20 28 25 13*

Gap

25

31*

27

* Significantly different (p<.05) from the nation (public) when comparing one state to the nation at a time.1 Department of Defense Education Activity (overseas and domestic schools).NOTE: States whose Hispanic student population size was insufficient for comparison are omitted. Reporting stan-dards not met for Maine, North Dakota, Vermont, and West Virginia. Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Reading Assessment.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


46

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Trends in state reading achievement gaps at grade 4The Hispanic-White reading gap among the nation’s public school fourth-graders did not change significantly, comparing 2009 to 1992, though both Hispanic and White students showed significant score increases (figure 26, national results).

From 2007 to 2009, scores of Hispanic and White fourth-graders in the nation did not change significantly, and there was also no significant change in the gap.

In 11 of the 21 states for which 1992 data were available, both Hispanic and White students achieved higher average scores in reading in 2009 than in 1992. Both New Jersey and New York narrowed the achievement gap when com-paring 2009 to 1992, as Hispanic students’ scores increased more than White students’ scores.

In Colorado, the gap was wider in 2009 than in 1992, as White students’ scores increased and Hispanic students’ scores showed no significant change. In Indiana, the gap was wider in 2009 than in 2002, the first assessment year for which reportable results for both groups are available, as Hispanic students’ scores decreased while White students’ scores did not change significantly.

From 2007 to 2009, scores increased for White students only in Rhode Island, and for Hispanic students only in Florida and Maryland. The gap did not change significant-ly in any of these states. In Kentucky, scores for White stu-dents increased from 2007 to 2009. Prior to 2009, Kentucky did not have reportable results for Hispanic students.

Stat

e n G

rade

4

Changing of the Gap

In New Jersey and New York, the gap was narrower in

Nar

row

ing 2009 than in the first assessment year as scores for

Hispanic students increased more than the scores for White students.

In Alaska, the gap was narrower in 2009 than in 2007, though neither Hispanic nor White students’ scores showed significant change.

In Colorado, the gap was wider in 2009 than in the first assessment year as scores for White students increased while scores for Hispanic students did not

iden

ing change.

In Indiana, the gap was wider in 2009 than in the first assessment year for which reliable results for W

both groups were available as Hispanic students’ scores decreased while White students’ scores did not change significantly.

In Alaska, the Hispanic-White reading gap narrowed between 2007 and 2009, even though scores for neither group changed significantly.

Figure 26 displays population percentages for Hispanic and White fourth-graders in each state for 2009 and the first year of the state’s participation in the assessment.

Hispanic-White reading gap data are not available for Maine, North Dakota, Vermont, or West Virginia.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Re

ad

ing

47

Figure 26. Reading achievement score gaps between Hispanic and White public school students at grade 4, by state: Various years, 1992–2009

Scale score

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 2003 2005 2007 2009

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

1992n 1994n 1998 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

1992n 1994n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1992n 1994n 1998

1992n 1994n

2002 2003 2005 2007 2009

229230228*227*227*223*223*222*223*

204204201*199*199*188*192*186*

194*

225227220*219*218*222219*217*

200197

226228225226

215206209209

225224224223220*219*219*220*

198197192*195188*188*188*197

224226225223222216*217*218*

202202212204204

227227225224223217*212*217*

196195193191*192181*

171*180*

236234232*232*226*220*221*

204204206205201191*

202

256258252254248247*248246*

207206193*187*193*

173*183*189*

234235232230*230*227*

223223219218*222212*

233232228*229*226*217*217*218*

223218*215*211*207*198*192*

203*

238238234*238237237233*230*

205203203206204196

183*187*

235233235233233218*215*221*

216218216209212

176*

National average

National average

National average

National average



Arkansas(2009: Hispanic 8%, White 66%)(1992: Hispanic #, White 75%)






DoDEA3



NATIONAL1


WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic

23* 1117 15

31 2732 2831 273123 24131918 323333 3135 3122

40*37

30 3227 26262920*3531333341 33

51*4324 18 181521

42

15 24* 20 19* 18* 13 14 1059 52 4955 67*

74*64*57

30 2528 26 26 2528*35*3136*28


12 11 1312815


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


48

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Stat

e n G

rade

4

Figure 26. Reading achievement score gaps between Hispanic and White public school students at grade 4, by state: Various years, 1992–2009—Continued

Scale score

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

1992n 2002 2003 2005 2007 2009 2003 2005 2007 2009 1992n 1994n 2002 2003 2005 2007 2009

1992n 1994n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

229230226226226221*221*223*

208212203201200

226227224221219214*214*212*

215205211204203197*189*193*

225227226222*224221*

201204199199197198

231230230228

203205199197

227226223*224225224*224

203207208212216*

224227224226225222224226

207208200205203

229229225225226227

210209203207205201

228225*222*221*222*220*214*214*

215

219220223*223*221218213*215

206213

225226225224225225229*227237236232231*230*224*222*220*

221213*210*209*208*207*197*

229230228*227*227*223*223*222*223*

204204201*199*199*188*192*186*

194*

National average

National average

National average

National average

NATIONAL1


WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic

28 26 26 2528*35*3136*28

27 27 23 242323











18 21232526 1922 121316 181725

2431 2831 181612* 249*

21 24 19 1622 20 1922182125 13

7 13


23 23 152218 2124



1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Re

ad

ing

49


Scale score

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1994n 1998 2002 2003 2005 2007 2009

1992n 1994n 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1992n 1994n 2003 2005 2007 2009

1992n 1994n 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

241241237*234*239228*230*230*

211209203*202*207194*

182*196*

225227226228*226223222

206210208205205201

230231231229229224*221*223*

194200204195202

225222220*221218*215*218*217*

212

228226226227226221*221*225*

216213210218

228230228227226227225

219220226

228230228225*226223*224*

207203202202203199205

222224217*218213*

199196

219

194192*195189*

230230228229224*228

217209206

237238232*235231*233*

213214206*212193*195*

224228225222223222220223

201204199197202195*197199

233234232235235228*226*226*

210206208208204188*189*184*

National average

National average

National average

National average


WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic

32 32 3035323447*34











17 1919242122 31 36263426

14 1212 9 16 101 9

26 27 2123232419 27 23242525 222023 13

24 24262437*38*


24 27 22273040*37*42*25 23 222521272323


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


50

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Stat

e n G

rade

4


Scale score

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 2002 2003 2005 2007 2009

1992n 2002 2003 2005 2007 2009 1992n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

1992n 1994n 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

229230228*227*227*223*223*222*223*

204204201*199*199*188*192*186*

194*

230228227232232223*224*220*

204205204212213*

228229228224*226227226

230231230226*229220*

215214211207

223223219220220225223

207198204200197204207

223222223222223217*

196190194199200

178*

230233229227228224*227

199200203195197191

231227*224*224*227*226*225*223*

200198192*196195177*

193183*

226224225226225221*218*221*

205205215

205

227228226227

216209

224224222220220218*219*218*

202208199206

192

232232232227*232230226*223*

210212210205*208200*198*200*

National average

National average

National average

National average

NATIONAL1


WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic

28 26 26 2528*

19 17 1519

35*3136*28


North Dakota(2009: Hispanic 2%, White 86%)(1992: Hispanic #, White 96%)








Tennessee(2009: Hispanic 5%, White 70%)(1992: Hispanic 1%, White 75%)

23 26232019

25 1615212316 21 3229 27232439

27 33 31323135 29 31322832

16 22221428

32 48*40199*21 22

19 11


22 21 22222430*2823



1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Re

ad

ing

51


Scale score

0

160

210

260

500

0

160

210

260

500

0

160

210

260

500

1992n 1994n 1998 2002 2003 2005 2007 2009 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009

1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2002 2003 2005 2007 2009 1992n 1994n 1998 2003 2005 2007 2009

1992n 1994n 1998 2002 2003 2005 2007 2009

225226226223224220*219*222*

194201199194201190192

200

229229227*226*227 234233233231233225*224*227*

214216218210

224*207211

229229228226227221*216*

201206202201204200185*

215216215220*220*216214216227229227225228227227

202208208209201203209

224228*227224224220*223225

212210204214207205208206

National average

National average

National average


WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

29 27 26 3123*2927

25 26 23 28232232

21

Vermont(2009: Hispanic 1%, White 94%)(2002: Hispanic 1%, White 95%)






13 17 2015219*18

2118 2516*272418

23 18 131115 171519




* Significantly different (p<.05) from 2009.1 National results for assessments prior to 2002 are based on the national sample, not on aggregated state samples.2 Hispanic and White percentages are based on students tested in 2009.3 Department of Defense Education Activity (overseas and domestic schools). Before 2005, DoDEA overseas and domestic schools were separate jurisdictions in NAEP. Pre-2005 data presented here were recalculated for comparability.NOTE: Score gaps are calculated based on differences between unrounded average scores. Where data are not present, the jurisdiction did not participate or did not meet the minimum partici-pation guidelines for reporting. State-level data were not collected in 2000. Comparative performance results may be affected by changes in exclusion rates for students with disabilities and English language learners in the NAEP samples.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1992–2009 Reading Assessments.

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


52

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Stat

e n G

rade

8

State and national reading achievement gaps at grade 8, 2009At grade 8, seven states had a smaller gap than the nation’s 24-point Hispanic-White gap in 2009 and none had a gap that was larger. In 36 states, the gap was not significantly different from the nation’s gap (figure 27). Asterisks indi-cate a significant difference from the national average score or the national average gap.

The eighth-grade reading gap in 2009 was statistically significant in 42 of the 43 states for which data could be

reported. In Kentucky, the apparent 3-point difference between the average scores for Hispanic and White stu-dents was not statistically significant, and thus there was no measurable gap.

In all seven of the states where the gap was smaller than the national gap, scores for Hispanic students were higher than the national average for those students. In DoDEA, scores for White students were also higher than the national aver-

1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


1990 1991 1992 1993 1994 1995 1996 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009I

II

II


Re

ad

ing

53

age for White students, while in Alaska, Florida, Missouri, and Wyoming, scores for White students were not measurably dif-ferent from the national average for those students. In Kentucky and South Carolina, scores for White students were lower than they were nationally.

Figure 27. The Hispanic-White achievement score gap in reading for public school students at grade 8, by state: 2009

5000

Scale score

230 240 250 260 270 280 290 300 310

Wyoming

Jurisdictions WhiteHispanic

WisconsinWashington

VirginiaUtah

TexasTennessee


OregonOklahoma

OhioNorth Carolina


New HampshireNevada

NebraskaMissouri

MinnesotaMichigan

MassachusettsMarylandKentucky

KansasIowa



DelawareConnecticut

ColoradoCaliforniaArkansas

ArizonaAlaska

Alabama

National Public 248

245260*

246249

241*250

252256*

269*260*

254252

241*252*251

249250

265*258*

250253

247260*

253* 242*

257256*

248247

249251

246247247

241*259*

252251*

246256*

248250

259*

271

264* 269 270

266* 269

274* 279*

273 278*

272 268*

267* 269*

274 269

267* 272

269* 279* 279*

268* 275*

270 272

264* 271

281* 271

275* 270

273 264*

269* 276*

267* 267* 267*

273 270

272 273

271 269

Gap

19 9* 24 17 28 24 27 16 9* 11* 14 15 28 21 18 18 22 3* 20 28 15 28 10* 19 22 14 25 24 27 22 22 18 22 28 26 8* 16 22 24 16 24 21 11*

24

* Significantly different (p<.05) from the nation (public) when comparing one state to the nation at a time.1 Department of Defense Education Activity (overseas and domestic schools).NOTE: States whose Hispanic or White student population size was insufficient for comparison are omitted. Reporting standards not met for District of Columbia, Lousiana, Maine, Mississippi, Montana, North Dakota, South Dakota, Vermont, and West Virginia. Score gaps are calculated based on differences between unrounded average scores.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Reading Assessment.

54

Stat

e n G

rade

8

Trends in state reading achievement gaps at grade 8The Hispanic-White reading gap among the nation’s public school eighth-graders did not change significantly, comparing 2009 to 1998, though both Hispanic and White students showed significant score increases (figure 28, national results).

From 2007 to 2009, scores of both Hispanic and White eighth-graders in the nation increased significantly. However, there was no significant change in the gap dur-ing this time period.

State-level data are available starting in 1998 for 22 states. In one state, Wyoming, both Hispanic and White students achieved higher average scores in reading in 2009 than in 1998. From 1998 to 2009, the Hispanic-White reading gap did not change significantly in any state (figure 28).

Scores increased for White students only in New Mexico and Utah, but the gap did not change significantly in either state. Scores increased for White students in Kentucky from 2007 to 2009; prior to 2009, Kentucky did not have reliable results for Hispanic students.

In Alaska, the gap narrowed in 2009 compared to 2003, the state’s first assessment year, as Hispanic students’ average scores increased while scores for White students showed no significant change.

Changing of the Gap

In Alaska, the gap was narrower in 2009 than in the first assessment year as Hispanic students’ average scores increased while scores for White students

Nar

row

ing showed no significant change.

In the following three states, the gap narrowed between 2007 and 2009 as Hispanic students’ aver-age scores increased while those of their White peers showed no change.

Rhode Island WyomingSouth Carolina

From 2007 to 2009, scores increased for Hispanic students only in Rhode Island, South Carolina, and Wyoming, and the gap narrowed in these three states during this period.

Figure 28 displays population percentages for Hispanic and White eighth-graders in each state for 2009 and the first year of the state’s participation in the assessment.

Hispanic-White reading gap data are not available for Louisiana, Maine, Mississippi, Montana, North Dakota, South Dakota, Vermont, or West Virginia.

Re

ad

ing

55

Figure 28. Reading achievement score gaps between Hispanic and White public school students at grade 8, by state: Various years, 1992–2009

Scale score

1992n 1994n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 2003 2005 2007 2009

1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

1998 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

265* 265* 268* 271 270* 269* 270* 271

238* 239* 241* 245 244* 245* 246* 248

265 264 262 263 261 264

250 245

268 268 270 269

246* 254 257 260

269 267 268 267 269 270

244 242 240* 242 241 246

263* 267 266 266 266 266

257 250 249 249

268 265 265 264* 266 269

238 238 237* 239* 239 241

270* 275 273 275 274

244 247 247 249 250

277 277275*272* 276* 279

247239*244*245 243*

252

263*275 273 274 274 273

248 250246* 253 257 256

301

246 240* 240* 247 249 249

276 278 278 276 278 278

269 270 268 268 273 269264* 269 268 265* 268 272

247 252*251*252* 256 260

National average

National average

National average

National average

NATIONAL1


WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic

26 24 25 2427*27

25 25 28 242825

2527

Alabama(2009: Hispanic 3%, White 60%)(1998: Hispanic 1%, White 63%)



Arkansas(2009: Hispanic 7%, White 69%)(1998: Hispanic 2%, White 75%)






DoDEA3


12 19 13 91421*

18 17169 2625 28292730

26 25 242726 33 27283138*30

5 99987

172127*25*15 16

53


13 12 11171717


56

Stat

e n G

rade

8


Scale score

1992n 1994n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

2002 2003 2005 2007 2009 2003 2005 2007 2009 2002 2003 2005 2007 2009

2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

265* 265* 268* 271 270* 269* 270* 271

238* 239* 241* 245 244* 245* 246* 248

Gap data not availableGap data not available

268 268 268 268 271 268

242 245 247 250 254

262 263 259* 261 262 267

246249 242 249 252

269267 267 268 269

247242 246 243 241

276 272 271 274

250 253 250 252

267 269 265* 268 269

247 247 255 251

269 269 270 267

244256 250 249

272 273271 271 272 272

241253

245 249 248 250

264* 267 269 266 264* 269

265

262 268* 267 264 264 263272* 270 269 270 270 268 272* 274271*272* 276 279

261 253 251 256 258 258

National average

National average

National average

National average

WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic











21 14202425 13 15201017

21 211926 1318 1822

13 19 1825 24 2222262031 3


16 18 20202111

NATIONAL1


26 24 25 2427*

21 20* 25 2825

272527


Re

ad

ing

57


Scale score

1998 2002 2003 2005 2007 2009 2002 2003 2005 2007 2009 1998 2003 2005 2007 2009

1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 2003 2005 2007 2009

2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

274* 278 278 279 278 279

242 246 246 246 251 250

270 272*268 267 268

257 250241

253

269* 273 273 273 275

240 244 245 247

264 268 267 264 264 264 265* 271 272 270 270 270

258248

260

273 273 273 272 274 273

273 271 271 271 272

251241* 245*

255 253

264 259* 262 261 263 264

242 237*237*241 238 242

272 270 270 271

252 257

277* 278 278 281

248* 251 257 256

270 266 268 264* 265* 271

250 247 243* 245 246 248

275 274 277 276 274 275

247 251 250 250 246 247

National average

National average

National average

National average

WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic











26 151816 28 282932

22 1012

26* 16 1930*2224 2219252222

20 2419252020

18 14

22 252628


26 29 27282328


31 32 27 283232

Gap data not availableGap data not available


58

Stat

e n G

rade

8


Scale score

1992n 1994n 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 2002 2003 2005 2007 2009

2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

265* 265* 268* 271 270* 269* 270* 271

238* 239* 241* 245 244* 245* 246* 248

270 274 271 267 270 270

252244 248 246 249

269* 272 272 270 271

273271 272 274 273

268*

245260

251

268* 268* 267 265 266 264

254 251 250 247 241 246

269 270 267 267 270 269

237249 249 245 243 247

271* 268* 273 272 276

241257

246 244 247

268 268267 268 267 267

239 240238 237 233*241

265 268269 267 268 267

244*259

273 272 272 273264 265265 265 267 267

252 252

271 276272 270 275 273

250 250247 248 251 251

National average

National average

National average

National average

WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

WhiteGapHispanic








South Carolina(2009: Hispanic 4%, White 55%)(1998: Hispanic 1%, White 58%)


Tennessee(2009: Hispanic 3%, White 70%)(1998: Hispanic 1%, White 76%)

24 22202722

25 1816 191714 223 27 14 2623 22172232

27 28 2811*31 34* 26

15 16

31302829 24* 8


22 24 22242622

NATIONAL1


26 24 25 2427*272527Gap data not available



Re

ad

ing

59


Scale score

1998 2002 2003 2005 2007 2009 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009

1998 2002 2003 2005 2007 2009 1998 2002 2003 2005 2007 2009 1998 2003 2005 2007 2009

1998 2002 2003 2005 2007 2009

0

200

250

300

500

0

200

250

300

500

0

200

250

300

500

266* 267 268 265* 266* 270

244 238* 241 243 242 246

272 271 269* 273 272 273 275 275 275 273 272

265 261 266 259 258 256

267* 271 268* 268* 270 273

240 247 246 245 247 248

262* 264*260*256 256 255269 271 271 270 271

256244 247 247 250

265* 267 269 270 269 269

250* 249* 255 256 248*259

National average

National average

National average

WhiteGapHispanic

WhiteGapHispanic

White

GapHispanic

Vermont(2009: Hispanic 1%, White 94%)(2002: Hispanic #, White 96%)






14 16169148

2224 212813

14 21* 111418*15


30 22 24 242721

24 23 23 242227Gap data not available



* Significantly different (p<.05) from 2009.1 National results for assessments prior to 2002 are based on the national sample, not on aggregated state samples.2 Hispanic and White percentages are based on students tested in 2009.3 Department of Defense Education Activity (overseas and domestic schools). Before 2005, DoDEA overseas and domestic schools were separate jurisdictions in NAEP. Pre-2005 data presented here were recalculated for comparability.NOTE: Score gaps are calculated based on differences between unrounded average scores. Where data are not present, the jurisdiction did not participate or did not meet the minimum partici-pation guidelines for reporting. State-level data were not collected in 1992, 1994, or 2000. Comparative performance results may be affected by changes in exclusion rates for students with disabilities and English language learners in the NAEP samples.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), various years, 1992–2009 Reading Assessments.

60

Reading Summary

National trends in reading, 1992–2009n The reading achievement gap between Hispanic and

White students in 2009 was 25 points at grade 4 and 24 points at grade 8. There was no significant difference when comparing these gaps to the comparable gaps in 2007 or 1992 (figures 17 and 18).

n Neither males nor females demonstrated a significant change in the Hispanic-White gap when comparing 2009 to 2007 or to 1992 (figures 19 and 20).

n At both the fourth and eighth grades, students who were eligible for free or reduced-price school lunch through the National School Lunch Program had a narrower Hispanic-White achievement gap in 2009 than in 2003 (figures 21 and 22).

n The gap between White and non-ELL Hispanic stu-dents at both grades was narrower in 2009 than in 1998, the first year for which data were available. The eighth-grade gap between Non-ELL Hispanic and ELL Hispanic students was wider in 2009 than in 2007 and 1998 (figures 23 and 24).

State achievement gaps in reading, compared to the nation, 2009n At grade 4, thirteen states had a smaller Hispanic-White

gap than the nation, and six had a gap that was larger (figure 25).

n At grade 8, seven states had a smaller Hispanic-White gap than the nation, and no state had a gap that was larger (figure 27).

State trends in reading, 1992–2009n At grade 4, of the 21 states for which 1992 data were

available, two states had a narrower Hispanic-White achievement gap in 2009 than in 1992, and one had a gap that was wider. One state had a narrower Hispanic-White achievement gap in 2009 than in 2007 (figure 26).

n At grade 8, three states had a narrower gap in 2009 than in 2007 (figure 28).

Re

ad

ing

61

62

ReferencesBarton, P., and Coley, R. (2007). The Family: America’s

Smallest School. Princeton, NJ: Educational Testing Service.

Barton, P., and Coley, R. (2009). Parsing the Achievement Gap II. Princeton, NJ: Educational Testing Service.

Campbell, J.R., Hombo, C.M., and Mazzeo, J. (2000). NAEP 1999 Trends in Academic Progress: Three Decades of Student Performance (NCES 2000-469). U.S. Department of Education, National Center for Education Statistics, Office of Educational Research and Improvement. Washington, DC.

Child Nutrition Program—Income Eligibility Guidelines. 73 Fed. Reg. 69 (April 9, 2008), pp. 19186-19187

Elementary and Secondary Education Act of 1965, P.L. 89-10, 79 Stat. 27.

Fry, R. (2007). How Far Behind in Math and Reading are English Language Learners? Washington, DC: Pew Hispanic Center.

Guzman, B. (2001). The Hispanic Population: Census 2000 Brief. (C2KBR/01-3), U.S. Census Bureau. Washington, DC.

Hill, H.C., Rowan, B., and Ball, D.L. (2005). Effects of Teachers’ Mathematical Knowledge for Teaching on Student Achievement. American Educational Research Journal, 42(2): 371-406.

Humes, K.R., Jones, N.A., and Ramirez, R.R. (2011). Overview of Race and Hispanic Origin: 2010 (C2010BR-02). Retrieved 29 March 2011, from U.S. Census Bureau website: http://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf

Johnson, E.G., and Rust, K.F. (1992). Population Inferences and Variance Estimation for NAEP Data. Journal of Educational Statistics, 17(2): 175-190.

KewalRamani, A., Fox, M., and Aud, S. (2010). Status and Trends in the Education of Racial and Ethnic Groups (NCES 2010-015). U.S. Department of Education, National Center for Education Statistics, Institute of Education Sciences. Washington, DC: U.S. Government Printing Office.

Lopez, M. (2009). Latinos and Education: Explaining the Attainment Gap. Washington, DC: Pew Hispanic Center.

Miller, R.G. (1981). Simultaneous Statistical Inference (2nd ed.). New York: Springer-Verlang.

National Center for Education Statistics (2009). The Nation’s Report Card: Mathematics 2009 (NCES 2010-451). Institute of Education Sciences, U.S. Department of Education, Washington, D.C.

National Center for Education Statistics (2010). The Nation’s Report Card: Reading 2009 (NCES 2010-458). Institute of Education Sciences, U.S. Department of Education, Washington, D.C.

National Research Council (2006). Hispanics and the Future of America. M. Tienda and F. Mitchell (Eds.), Committee on Population, Division of Behavioral and Social Sciences and Education. Washington, DC: The National Academies Press.

U.S. Census Bureau, 2010 Census. (2010). Race, Hispanic or Latino, Age, and Housing Occupancy: 2010. 2010 Census Redistricting Data (Public Law 94-171) Summary File, Tables P1, P2, P3, P4, H1. Retrieved 29 March 2011, from U.S. Census Bureau website: http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=DEC_10_PL_QTPL&prodType=table

Vanneman, A., Hamilton, L., Baldwin Anderson, J., and Rahman, T. (2009). Achievement Gaps: How Black and White Students in Public Schools Perform in Mathematics and Reading on the National Assessment of Educational Progress (NCES 2009-455). U.S. Department of Education, National Center for Education Statistics, Institute of Education Sciences. Washington, DC.

Williams, V.S.L., Jones, L.V., and Tukey, J.W. (1994). Controlling Error in Multiple Comparisons with Special Attention to the National Assessment of Educational Progress. Research Triangle Park, NC: National Institute of Statistical Sciences.

http://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf

http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=DEC_10_PL_QTPL&prodType=table

64

Appendix A: Technical NotesThis report presents data for public school students from the main National Assessment of Educational Progress (NAEP). Main NAEP assessments are conducted in a range of subjects at grades 4, 8, and 12 across the country, includ-ing the District of Columbia and Department of Defense Education Activity (DoDEA), and are used to report at the national, state, and district levels. Main NAEP assessments began in 1990 for mathematics and 1992 for reading at grades 4 and 8. Discussion of main NAEP grade 12 assess-ments is omitted in this report because these assessments are conducted at the national level only.

Frameworks, development, administration, scoring, and analysisFor overviews of these topics, and for more extensive infor-mation about other topics for the 2009 main NAEP reading and mathematics assessments, consult the information avail-able online at http://nces.ed.gov/nationsreportcard/reading/ and http://nces.ed.gov/nationsreportcard/mathematics/.

Sources of the dataThis report presents national data from the 1990, 1992, 1996, 2000, 2003, 2005, 2007, and 2009 main NAEP math-ematics assessments and the 1992, 1994, 1998, 2002, 2003, 2005, 2007, and 2009 main NAEP reading assessments for White and Hispanic public school students in the fourth and eighth grades. In 2000, the reading assessment was administered at the fourth grade only (see tables B-1 and B-3 in appendix B).

This report presents state data from the 1992, 1996, 2000, 2003, 2005, 2007, and 2009 grade 4 NAEP mathematics assessments and from the 1990, 1992, 1996, 2000, 2003, 2005, 2007, and 2009 grade 8 NAEP mathematics assess-ments, for public school students only. It presents state data from the 1992, 1994, 1998, 2002, 2003, 2005, 2007, and 2009 grade 4 reading assessments and from the 1998, 2002, 2003, 2005, 2007, and 2009 grade 8 reading assessments.

Nationally in 2009, White students constituted 54 percent of the public school fourth-grade population while Hispanic students constituted 22 percent, using data from the 2009 mathematics assessment. Results for the eighth-grade were

similar: 56 percent and 21 percent, respectively. However, percentages vary widely across states. For example, Hispanic students constituted a majority of the fourth-grade popula-tion in three states, California (51 percent), Texas (51 per-cent), and New Mexico (58 percent), according to the 2009 mathematics assessment. In contrast, Hispanic students constituted 1 percent of the fourth-grade public school popu-lation in Maine, Vermont, and West Virginia. Eighth-grade data show a similar pattern. In some states, the NAEP sam-ple for the Hispanic or White population was insufficient to permit a reliable estimate.

NAEP sampling procedures for public school students in reading and mathematicsThe schools and students participating in NAEP assess-ments are chosen to be representative of the nation and states. Samples of public schools and students are selected from each state and from the District of Columbia and DoDEA schools. The results from the assessed students are combined to provide accurate estimates of overall national performance and of the performance of individual states.

The National Center for Education Statistics (NCES) has changed the main NAEP sampling methods over the years. From 1990 through 2000, the national public sample was collected separately from the state samples. The 2002 national sample was the sum of all the state samples of the participating states, plus small samples from the few states that did not participate. In 2003, 2005, 2007, and 2009, all states participated and the national sample was the aggregate of the samples from all states and the District of Columbia. (As discussed below, DoDEA schools are not considered public schools, although for comparison purposes DoDEA is treated as a state.) The main NAEP national samples in reading and mathematics since 2002 have been larger than in previous assessment years. Thus, smaller score dif-ferences between years or between types of student groups were found to be statistically significant than would have been detected in previous assessments. From 1990 through 2001, NCES oversampled schools with high minority popu-lations (Black and Hispanic) in the national public sample. Beginning in 2002, this practice was discontinued because

65

the state samples were large enough to ensure adequate coverage for these populations. Prior to 2002, NAEP results were weighted to compensate for the oversampling.

In 2002, 2003, 2005, 2007, and 2009, results were weighted to take into account the fact that states, and schools within states, represent different proportions of the overall nation-al public population. For example, since the number of stu-dents assessed in most states is roughly the same (to allow for stable state estimates and administrative efficiencies), the results for students in less populous states are assigned smaller weights than the results for students in more popu-lous states. Sampling weights are also used to account for lower sampling rates for very small schools and are used to adjust for school and student nonresponse. NAEP samples for reading and mathematics assessments administered from 1990 through 2009 are discussed in more detail below.

The NAEP 2009 mathematics and reading assessments were administered to fourth- and eighth-graders in all states. This report includes data for public school students for both the nation and all states. All 50 states, the District of Columbia, and the DoDEA schools met the minimum guidelines for reporting their results in 2009 for both assessments.

In order to obtain a representative sample for reporting national and state public school results in 2009, NCES sampled and assessed approximately 168,800 fourth-graders from 9,510 schools and 161,700 eighth-graders from 7,030 schools for the mathematics assessment and approximately 178,800 fourth-graders from 9,530 schools and 160,900 eighth-graders from 7,030 schools for the reading assessment.

Each student assessed represented a portion of the popula-tion. The students selected represented the total popula-tion of approximately 3.5 million fourth-grade and 3.5 million eighth-grade public school students. These totals include the public schools in the 50 states and the District of Columbia.

Schools in the DoDEA school system are classified as “nonpublic” by NCES and their results are not included in the determination of NAEP national public average scale scores. These schools are not “private” because they are operated by the federal government and they are not

“public” because only children of U.S. military personnel can attend them. For comparison purposes, the system is treated as a state and results are compared with the scores of the 50 states and the District of Columbia.

School and student participationTable A-1 provides a summary of the 2009 national and state school and student participation rates for the reading grade 8 assessment sample. Rates for reading grades 4 and 8 and mathematics grade 4 in 2009 were similar, as were the rates for the 2002, 2003, 2005, and 2007 assessments. Readers who want more detail should consult the 2009, 2007, 2005, 2003 and 2002 report cards, available online at http://nces.ed.gov/pubsearch/.

Participation rates in table A-1 are presented for public schools and public school students in grade 8 reading. The school participation rate is a weighted percentage of schools participating in the assessment. This rate is based only on the schools that were initially selected for the assessment, before substitution. The numerator of this rate is the estimated number of schools represented by the initially selected schools that participated in the assessment. The denominator is the estimated number of schools rep-resented by the initially selected schools that had eligible students enrolled.

Also presented in table A-1 are weighted student par-ticipation rates. The numerator of this rate is the estimated number of students who are represented by the students assessed (in either an initial session or a makeup session). The denominator of this rate is the estimated number of students represented by the eligible sampled students in participating schools.

The term “eligible students” used in the two preceding paragraphs refers to students who can meaningfully partici-pate in NAEP. Students excluded from NAEP assessments on the grounds that they cannot meaningfully participate are not part of the population of interest. Initially selected schools that had no eligible students enrolled are excluded from the denominator of the school participation rate because they contained no students who were part of the population of interest. For similar reasons, the denominator

66

Table A-1. School and student participation rates, and target student population, grade 8 reading assessment, public school students only, by state or jurisdiction: 2009

School participation Student participation

JurisdictionSchool-weighted

percentageNumber of schools

participatingStudent-weighted

percentageNumber of students

assessed Target populationNational Public 100 6,510 92 155,400 3,504,000

Alabama 100 110 93 2,700 53,000Alaska 88 100 91 2,500 9,000Arizona 100 130 92 2,800 73,000Arkansas 100 120 93 2,700 33,000California 100 230 92 7,200 469,000Colorado 100 120 93 2,800 54,000Connecticut 100 110 92 2,800 42,000Delaware 100 50 92 2,800 9,000District of Columbia 100 60 89 1,600 4,000DoDEA1 97 60 93 1,600 5,000Florida 100 160 91 4,200 180,000Georgia 100 120 93 3,500 109,000Hawaii 100 70 92 2,900 13,000Idaho 100 110 94 3,000 20,000Illinois 100 200 94 4,100 154,000Indiana 100 110 93 2,700 77,000Iowa 100 130 94 2,600 33,000Kansas 99 120 95 2,700 33,000Kentucky 100 130 94 3,500 47,000Louisiana 100 120 93 2,600 45,000Maine 100 140 93 2,700 14,000Maryland 100 130 92 3,200 58,000Massachusetts 100 140 92 3,600 72,000Michigan 100 150 92 3,300 117,000Minnesota 100 140 92 2,900 60,000Mississippi 100 120 94 2,800 37,000Missouri 100 130 94 2,700 64,000Montana 98 170 91 2,600 11,000Nebraska 100 120 95 2,600 20,000Nevada 100 90 92 2,900 32,000New Hampshire 96 90 90 2,500 15,000New Jersey 100 110 93 2,700 100,000New Mexico 100 100 90 2,500 23,000New York 98 150 90 3,700 198,000North Carolina 100 150 92 4,500 112,000North Dakota 100 180 95 2,100 7,000Ohio 100 190 93 3,400 129,000Oklahoma 100 150 93 2,700 44,000Oregon 100 130 92 2,900 42,000Pennsylvania 100 150 92 3,500 127,000Rhode Island 100 60 92 2,700 11,000South Carolina 100 110 93 2,700 50,000South Dakota 100 220 95 2,800 9,000Tennessee 100 120 93 2,800 75,000Texas 100 170 92 5,700 322,000Utah 100 110 91 2,800 38,000Vermont 100 120 93 2,900 7,000Virginia 100 110 93 2,800 91,000Washington 100 130 91 2,800 75,000West Virginia 100 120 92 2,900 23,000Wisconsin 99 170 93 3,400 61,000Wyoming 100 90 91 1,900 6,0001 Department of Defense Education Activity (overseas and domestic schools).NOTE: The numbers of schools are rounded to the nearest ten, the numbers of students are rounded to the nearest hundred, and the target population is rounded to the nearest thousand. Detail may not sum to totals because of rounding.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Reading Assessment.

Appe

ndix

A: T

echn

ical

Not

es

67

of the weighted student participation rate consists only of eligible sampled students.

The fourth column gives the number of public school stu-dents who were assessed in each of the jurisdictions. The final column of table A-1 gives the target populations for each jurisdiction, that is, the eighth-grade population for that jurisdiction.

The national target student population per grade for all main NAEP assessments 1990–2009 ranged from about 3.25 million to about 3.75 million. In the 1990–1996 assess-ments, the number of schools sampled per assessment and grade for the national sample ranged from approximately 120 to 230, while the number of students assessed ranged from approximately 5,200 to 9,900. In the 1998–2000 assess-ments, the number of schools sampled per assessment and grade ranged from approximately 330 to 390, while the number of students assessed ranged from approximately 6,100 to 9,000. (Data are drawn from NAEP mathematics assessments, 1990–2009.)

The state target student populations for all main NAEP assessments 1990–2009 ranged from approximately 5,000 in the District of Columbia and 9,000 in sparsely populated states like Wyoming and Alaska to approximately 450,000 in California, followed by approximately 325,000 in Texas. In the 1990–2000 state assessments, the number of schools sampled per assessment and grade ranged from approxi-mately 30 to 150, while the number of students assessed ranged from approximately 1,000 to 5,900. In the 2002– 2009 state assessments, the number of schools sampled per assessment and grade ranged from approximately 40 to 250, while the number of students assessed ranged from approximately 1,700 to 10,700. (Data are drawn from NAEP mathematics assessments, 1990–2009.)

In all NAEP assessment years prior to 2003, NCES pre-selected substitute schools that could be added to the original sample in case a large number of schools from the sample failed to participate. School and student participa-tion rates were given both before and after substitution. Because the No Child Left Behind Act of 2001 requires states to participate in the main NAEP reading and mathematics assessments at the fourth and eighth grades

in order to qualify for full Title I education funding, par-ticipation rates are very high and NCES no longer selects substitute schools for these assessments.

NCES and the National Assessment Governing Board, which establishes policy for NAEP, set minimums for the school participation rate before substitution of replace-ment schools for any sample. Results are not reported for states with a participation rate below the minimum. From 1990 through 2002, the standard for the state assessments required that the weighted school participation rate before substitution of replacement schools be 70 percent or higher. Beginning in 2003, the standard was raised to 85 percent. All data presented in this report are based on samples meeting the standards in effect at the time of the assess-ment. Since 2003, no state has had a rate below 85 percent.

Since 1990, the national weighted public school participa-tion rate before substitution for the grade 4 and 8 reading and mathematics assessments has ranged from 76 percent to 100 percent. Prior to 2003, a few states did not meet the 70 percent standard. From 1990 through 2002, the weighted public school participation rate before substi-tution for states whose results are reported here ranged from 70 percent to 100 percent. For more information on all the NAEP assessments referenced in this report, consult the individual reports devoted to them, available from the NCES website at http://nces.ed.gov/pubsearch/getpubcats.asp?sid=031.

Understanding NAEP reporting groupsNAEP results are provided for groups of students defined by shared characteristics—race/ethnicity, eligibility for free or reduced-price school lunch, and gender, for example. Based on participation rate criteria, results are reported for groups only when sufficient numbers of students and adequate school representation are present. The minimum requirement is a total of at least 62 students in a particular group, assessed in at least five different locations. However, the data for all students, regardless of whether their group is reported separately, are included in computing overall stu-dent results. Definitions of the student groups discussed in

http://nces.ed.gov/pubsearch/getpubcats.asp?sid=031

68

this report follow. For more information on understanding NAEP reporting, see http://nces.ed.gov/nationsreportcard/mathematics/interpret-results.asp.

Race/ethnicityIn all main NAEP assessments, data about student race/ ethnicity are collected from two sources: school records and student self-reports. In this report, the race/ethnicity variable is based on the race reported by the school for all assessment years. In the rare cases when school-recorded information is missing, student-reported data are used to determine race/ethnicity.

Schools sampled for NAEP are asked to provide lists of all students in the target grade(s) along with basic demo-graphic information, including race/ethnicity. Students are categorized into one of five mutually exclusive racial/ ethnic categories plus “other.” Administration schedules—also referred to as student rosters—are created that include the list of sampled students along with their basic demo-graphic information. These data are checked and updated during data collection. These race/ethnicity data were col-lected for all sampled students: those that participated and those that were absent or excluded. See http://nces.ed.gov/nationsreportcard/bgquest.asp for more information.

The mutually exclusive racial/ethnic categories are White (non-Hispanic), Black (non-Hispanic), Hispanic, Asian/ Pacific Islander, American Indian (including Alaska Native), and Unclassified. Unclassified students are those whose school-reported race was “other” or “unavailable,” or was missing, or whose race could not be determined using self-reported data (i.e., “multi-racial” or missing). Hispanic students may be of any race. Only results for White (non-Hispanic) and Hispanic students are contained in this report. Information based on student self-reported race/ethnicity is available on the NAEP Data Explorer (http://nces.ed.gov/nationsreportcard/nde).

National School Lunch ProgramNAEP first began collecting data in 1996 on student eligi-bility for the National School Lunch Program (NSLP) as an indicator of low income. Under the guidelines of NSLP,

children from families with incomes below 130 percent of the poverty level are eligible for free meals. Those from families with incomes between 130 and 185 percent of the poverty level are eligible for reduced-price meals. (For the period July 1, 2008 through June 30, 2009, for a family of four, 130 percent of the poverty level was $27,560, and 185 percent was $39,220 [Child Nutrition Program 2008].)

Some schools provide free meals to all students irrespec-tive of individual eligibility, using their own funds to cover the costs of non-eligible students. Under special provisions of the National School Lunch Act intended to reduce the administrative burden of determining student eligibility every year, schools can be reimbursed based on eligibility data for a single base year. Participating schools might have high percentages of eligible students and report all students as eligible for free lunch. For more information on NSLP, visit http://www.fns.usda.gov/cnd/lunch/.

Because of the improved quality of the data on students’ eligibility for NSLP, the percentage of students for whom information was not available has decreased compared to the percentages reported prior to the 2003 assessment. Therefore, NSLP trend comparisons are only made back to 2003 in this report.

Gender NAEP assessments identify students as male or female based on school records.

Inclusion and exclusion The NAEP program has always endeavored to assess all students selected as a part of its sampling process. In all NAEP schools, accommodations are provided as necessary for students with disabilities (SD) and/or English lan-guage learner (ELL) or limited English proficient (LEP) students. (ELL is the term used since the NAEP 2005 reports; LEP was used before 2005.) The accommodations are available to students whose Individualized Education Program (IEP) specifically requires them. Because some ELL students do not have an IEP, decisions about accom-modations for these students are typically made by knowl-edgeable school staff.

Appe

ndix

A: T

echn

ical

Not

es

http://nces.ed.gov/nationsreportcard/mathematics/interpret-results.asp

http://nces.ed.gov/nationsreportcard/bgquest.asp

69

The NAEP program has established procedures to include as many SD and ELL students as possible in the assess-ments. School staff make the decisions about whether to include such a student in a NAEP assessment, and which testing accommodations, if any, they should receive. The NAEP program furnishes tools to assist school personnel in making those decisions.

A sampling procedure is used to select students at each grade being tested. Students are selected on a random basis, without regard to SD or ELL status. Once the students are selected, the schools identify which have SD or ELL status. School staff who are familiar with these students are asked a series of questions to help them decide whether each stu-dent should participate in the assessment and whether the student needs accommodations.

Inclusion in NAEP of an SD or ELL student is encour-aged if:

(a) that student participated in the regular state academic assessment in the subject being tested, and

(b) that student can participate in NAEP with the accom-modations NAEP allows.

Even if the student did not participate in the regular state assessment, or if he/she needs accommodations NAEP does not allow, school staff are asked whether that student could participate in NAEP with the allowable accom-modations. For more information on inclusion, exclusion, and accommodations in NAEP, visit http://nces.ed.gov/nationsreportcard/about/inclusion.asp.

History of NAEP Inclusion PolicyAlthough NAEP has always endeavored to assess as high a proportion of sampled students as is possible, prior to 1996 NAEP did not allow accommodations for SD or ELL students. This resulted in exclusion of some students who could not meaningfully participate in the assessment with-out accommodations.

The passage of the Individuals with Disabilities Education Act (IDEA), as amended in 1997, led states and districts to identify increasing numbers of students as requiring accommodations in assessments in order to fairly and

accurately show their abilities. It was important for NAEP to be as consistent as possible with testing practices in most states and districts while maintaining the ability to compare more recent NAEP results to those from 1990, 1992, and 1994, when accommodations were not allowed. (Accommodations were not allowed in NAEP state assess-ments until 1996.) Before the 2005 assessment (when the selection process was detailed in a series of questions), guidelines were specified by NAEP. Beginning in 2005, a student identified on the Administration Schedule as having a disability (SD), that is, a student with an IEP or equivalent classification, should be included in the NAEP assessment unless:

n The IEP team or equivalent group had determined that the student could not participate in assessments such as NAEP,

n The student’s cognitive functioning was so severely impaired that he or she could not participate, or

n The student’s IEP required that the student be tested with an accommodation that NAEP did not permit, and the student could not demonstrate his or her knowledge of the subject without that accommodation.

A student who was identified as LEP or ELL and who was a native speaker of a language other than English should be included in the NAEP assessment unless:

n The student had received reading or mathematics instruction primarily in English for less than 3 school years including the current year, and

n The student could not demonstrate his or her knowl-edge of the subject in English even with an accommoda-tion permitted by NAEP.

The phrase “less than 3 school years including the current year” meant 0, 1, or 2 school years. Therefore, the guide-lines below were used:

n Include without any accommodation all LEP or ELL students who had received instruction in the subject primarily in English for 3 years or more and those who were in their third year;

http://nces.ed.gov/nationsreportcard/about/inclusion.asp

70

n Include without any accommodation all other such students who could demonstrate their knowledge of the subject without an accommodation;

n Include and provide accommodations permitted by NAEP to other such students who can demonstrate their knowledge of the subject only with those accom-modations; and

n Exclude LEP or ELL students only if they could not demonstrate their knowledge of the subject even with an accommodation permitted by NAEP.

The percentages of students excluded from NAEP may vary from one state to another, as well as across years. National exclusion rates for Hispanic and White SD and/ or ELL students in 2009 may be found in table A-2. The “total” rates include all students, not just those who are Hispanic or White. For information on state exclu-sion rates, see table A-3. For more information on Main NAEP inclusion and exclusion, go to http://nces.ed.gov/nationsreportcard/about/inclusion.asp

Accommodations From 1990 through 1994 for the nation—and through 1996 for the states—main NAEP assessments did not allow accommodations for either SD or ELL students. Since then, accommodations have been permitted for those SD and/or ELL students who need accommodations in order to participate, unless the accommodation would change the nature of what is being tested.

To accomplish this goal, students who receive accommo-dations in their state’s assessments are offered the same accommodations on NAEP. For example, passages and questions in the NAEP reading test are not permitted to be read aloud to the student, because that accommodation would make it a test of listening instead of a test of read-ing. Similarly, reading passages and questions cannot be presented in a language other than English. It should be noted that students assessed with accommodations typi-cally received some combination of accommodations. For example, students assessed in small groups (as compared with standard NAEP sessions of about 30 students) usually received extended time. In one-on-one administrations,

students often received assistance in recording answers (e.g., use of a scribe or computer) and extra time.

The most common accommodations for the NAEP 2009 reading and mathematics assessments were small-group administration, extended time, breaks, and read-aloud (mathematics only). See http://nces.ed.gov/nationsreportcard/tdw/instruments/accomm.asp for more details on NAEP accommodations. For state accommodation rates for SD and ELL students in 2009 see the Technical Notes sec-tions of The Nation’s Report Card: Mathematics 2009 at http://nationsreportcard.gov/math_2009/ and The Nation’s Report Card: Reading 2009 at http://nationsreportcard.gov/reading_2009/.

Drawing inferences from the results The reported statistics for NAEP are estimates and there-fore subject to a measure of uncertainty. There are two sources of such uncertainty. First, NAEP uses a sample of students rather than testing all students. Second, all assess-ments have some amount of uncertainty related to the fact that they cannot ask all questions that might be asked in a content area. The magnitude of this uncertainty is reflected in the standard error of each of the estimates. When the percentages or average scale scores of certain groups are compared, the estimated standard error should be taken

Table A-2. National mathematics and reading exclusion rates as percentages of the total sample, public schools only, by grade and race/ethnicity: 2009

Math ReadingGrade 4

Total 2 5

White 2 3

Hispanic 3 8

Grade 8

Total 3 4

White 3 3

Hispanic 3 5

NOTE: “Total” exclusion percentages are for all public school students, not just Hispanic and White.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2007 Mathematics and Reading Assessments.

Appe

ndix

A: T

echn

ical

Not

es

http://nces.ed.gov/nationsreportcard/about/inclusion.asp

http://nces.ed.gov/nationsreportcard/tdw/instruments/accomm.asp

http://nationsreportcard.gov/reading_2009

71

Table A-3. Mathematics and reading exclusion rates as percentages of the total sample, public schools only, by grade, race/ethnicity, and jurisdiction: 2009

Mathematics Reading Grade 4 Grade 8 Grade 4 Grade 8Jurisdiction Hispanic White Hispanic White Hispanic White Hispanic WhiteNational Public 3 2 3 3 8 3 5 3

Alabama 2 1 6 1 5 1 9 1Alaska 3 1 4 3 3 2 2 1Arizona 1 1 2 2 5 3 3 2Arkansas 2 1 2 1 1 1 3 1California 3 1 2 1 3 3 2 1Colorado 1 2 3 1 4 3 4 3Connecticut 5 2 4 2 12 3 11 2Delaware 3 3 5 2 8 7 11 3District of Columbia 9 1 13 4 15 6 19 6DoDEA1 4 2 5 1 8 5 6 3Florida 3 2 3 2 10 2 8 2Georgia 1 1 3 2 12 4 9 3Hawaii 3 1 2 2 4 1 4 3Idaho 2 1 1 1 6 3 5 2Illinois 5 2 4 3 8 2 6 3Indiana 1 2 5 4 10 4 11 4Iowa 2 2 6 2 9 4 8 4Kansas 6 2 4 2 12 4 12 3Kentucky 5 3 7 4 20 7 16 7Louisiana 1 1 ‡ 1 5 2 ‡ 2Maine ‡ 2 ‡ 2 ‡ 4 ‡ 3Maryland 8 3 11 4 23 8 27 5Massachusetts 9 4 6 5 11 3 15 4Michigan 4 3 3 3 6 3 4 4Minnesota 4 2 4 2 6 2 6 3Mississippi ‡ 1 ‡ 1 5 1 ‡ 1Missouri 5 3 8 3 8 3 7 3Montana 0 1 9 3 4 3 ‡ 3Nebraska 3 2 5 3 9 4 11 5Nevada 3 2 3 2 6 3 5 2New Hampshire 2 2 6 3 11 3 2 4New Jersey 6 2 3 2 18 7 15 5New Mexico 2 3 3 2 10 4 7 3New York 1 1 2 2 8 3 10 5North Carolina 2 2 3 1 6 1 6 1North Dakota ‡ 4 ‡ 5 ‡ 7 ‡ 7Ohio 10 2 16 4 17 5 11 6Oklahoma 3 3 4 7 11 6 9 4Oregon 4 2 4 2 7 3 4 2Pennsylvania 5 2 7 2 9 2 6 2Rhode Island 3 1 5 2 7 3 7 2South Carolina 4 1 4 3 16 4 22 4South Dakota 3 2 1 2 11 5 ‡ 3Tennessee 3 3 10 4 14 8 15 6Texas 3 2 4 5 13 5 5 4Utah 4 2 3 3 14 5 9 3Vermont ‡ 2 ‡ 2 ‡ 3 ‡ 3Virginia 2 2 6 3 9 3 13 3Washington 2 1 2 2 5 3 3 3West Virginia ‡ 2 ‡ 2 ‡ 2 ‡ 2Wisconsin 5 1 4 2 7 3 11 3Wyoming 2 1 5 1 6 2 5 3

‡ Reporting standards not met.1 Department of Defense Education Activity (overseas and domestic schools). SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Mathematics and Reading Assessments.

72

into account. Therefore, the comparisons are based on statistical tests that consider the estimated standard errors of the statistics being compared and the magnitude of the difference between the averages or percentages. Estimates based on smaller groups are likely to have relatively large standard errors. As a consequence, a numerical difference that seems large may not be statistically significant.

Furthermore, differences of the same magnitude may or may not be statistically significant, depending upon the size of the standard errors of the statistics. For example, a 3-point change in the gap between Hispanic and White fourth-graders nationwide may be significant, while a 3- point change in the gap between Hispanic and White fourth-graders in Kansas may not be. The differences described in this report have been determined to be statis-tically significant at the .05 level with appropriate adjust-ments for part-to-whole and multiple comparisons.

In the tables and figures of this report, the symbol (*) is used to indicate that a score or percentage is significantly different from another. In addition, any difference between scores or percentages that is identified as higher, lower, larger, smaller, narrower, or wider in this report, including within-group differences not marked in tables and figures, meets the requirements for statistical significance.

Standard errors for the NAEP scores and percentages pre-sented in this report for both assessments are available on the NAEP website (http://nces.ed.gov/nationsreportcard/naepdata).

Weighting and variance estimation NAEP uses a complex sample design to select the students who were assessed. The properties of a sample selected through such a design could be very different from those of a simple random sample, in which every student in the target population has an equal chance of selection and in which the observations from different sampled students can be considered to be statistically independent of one another. Therefore, the properties of the sample for the data collection design were taken into account during the analysis of the assessment data.

One way that the properties of the sample design were addressed was by using sampling weights to account for the fact that the probabilities of selection were not iden-tical for all students. All population and subpopulation characteristics based on the assessment data were esti-mated using sampling weights. These weights included adjustments for school and student nonresponse.

Not only must appropriate estimates of population charac-teristics be derived, but appropriate measures of the degree of uncertainty must be obtained for those statistics. Two components of uncertainty are accounted for in the vari-ability of statistics based on student ability: (1) the uncer-tainty due to sampling only a relatively small number of students, and (2) the uncertainty due to sampling only a relatively small number of cognitive questions

Because NAEP uses complex sampling procedures, con-ventional formulas for estimating sampling variability that assume simple random sampling are inappropriate. NAEP uses a jackknife replication procedure to estimate standard errors. The jackknife standard error provides a reasonable measure of uncertainty for any student infor-mation that can be observed without error. However, because each student typically responds to only a few questions within a content area, the scale score for any single student would be imprecise. In this case, NAEP’s marginal estimation methodology can be used to describe the performance of groups and subgroups of students. The estimate of the variance of the students’ posterior scale score distributions (which reflect the imprecision due to lack of measurement accuracy) is computed. This component of variability is then included in the standard errors of NAEP scale scores.1

Analyzing group differences in averages and percentages NAEP uses statistical tests to determine whether, based on the data from the groups in the sample, there is strong enough evidence to conclude that the averages or per-

1 For further detail, see Johnson, E.G., and Rust, K.F. (1992). Population Inferences and Variance Estimation for NAEP Data. Journal of Educational Statistics, (17)2, 175–190.

Appe

ndix

A: T

echn

ical

Not

es

73

centages are actually different for those groups in the population. If the evidence is strong (i.e., the difference is statistically significant), the report describes the group averages or percentages as being different (e.g., one group performed higher or lower than another group), regardless of whether the sample averages or percentages appear to be approximately the same. The reader is cautioned to rely on the results of the statistical tests rather than on the apparent magnitude of the difference between sample averages or percentages when determining whether the sample differ-ences are likely to represent actual differences among the groups in the population.

To determine whether a real difference exists between the average scale scores (or percentages of a certain attribute) for two groups in the population, one needs to obtain an estimate of the degree of uncertainty associated with the difference between the averages (or percentages) of these groups for the sample. This estimate of the degree of uncertainty, called the “standard error of the difference” between the groups, is obtained by taking the square of each group’s standard error, summing the squared stan-dard errors, and taking the square root of that sum.

The standard error of the difference can be used, just like the standard error for an individual group average or percentage, to help determine whether differences among groups in the population are real. The difference between the averages or percentages of the two groups plus or minus 1.96 standard errors of the difference represents an approximately 95 percent confidence interval for a two-tailed test. If the resulting interval includes zero, there is insufficient evidence to claim a real difference between the groups in the population. If the interval does not contain zero, the difference between the groups is statistically sig-nificant at the .05 level.

The following example of comparing groups addresses the problem of determining whether the average mathematics scale score of group A is higher than that of group B. The sample estimates of the average scale scores and estimated standard errors are as follows:

Group Average scale score Standard error

A 218 0.9

B 216 1.1

The difference between the estimates of the average scale scores of groups A and B is 2 points (218 – 216). The stan-dard error of this difference is

Thus, an approximately 95 percent confidence interval for this difference is plus or minus 1.96 standard errors of the difference:

The value zero is within the confidence interval; therefore, there is insufficient evidence to conclude that group A’s performance is statistically different from group B.

The procedure above is appropriate to use when it is rea-sonable to assume that the groups being compared have been independently sampled for the assessment.

Such an assumption is clearly warranted when comparing results for one state with another. This is the approach used for NAEP reports when comparisons involving independent groups are made. The assumption of inde-pendence is violated to some degree when comparing group results for the nation or a particular state (e.g., comparing national 2009 results for male and female stu-dents), since these samples of students have been drawn from the same schools.

When the groups being compared do not share students (as is the case, for example, of comparing Hispanic and White students), the impact of this violation of the inde-pendence assumption on the outcome of the statistical tests is assumed to be small, and NAEP, by convention, has, for computational convenience, routinely applied the proce-dures described above to those cases as well.

When making comparisons of results for groups that share a considerable proportion of students in common, it is not appropriate to ignore such dependencies. In such cases,

74

NAEP has used procedures appropriate to comparing dependent groups. When the dependence in group results is due to the overlap in samples (e.g., when a subgroup is being compared to a total group), a simple modification of the usual standard error of the difference formula can be used. The formula for such cases is

where p is the proportion of the total group contained in the subgroup. This formula was used for this report when a state was compared to the aggregate for the nation.

Conducting multiple tests The procedures used to determine whether group dif-ferences in the NAEP samples represent actual differ-ences among the groups in the population and the certainty ascribed to intervals (e.g., a 95 percent confidence interval) are based on statistical theory that assumes that only one confidence interval or test of statistical significance is being performed. However, there are times when many different groups are being compared (i.e., multiple sets of confidence intervals are being analyzed).

For multiple comparisons, statistical theory indicates that the certainty associated with the entire set of compari-sons is less than that attributable to each individual com-parison from the set. To hold the significance level for the set of comparisons at a particular level (e.g., .05), the standard methods must be adjusted by multiple compari-son procedures (Miller 1981). In NAEP, the Benjamini-Hochberg False Discovery Rate (FDR) procedure is used to control the expected proportion of falsely rejected hypotheses relative to the number of comparisons that are conducted. A detailed explanation of this procedure

can be found at http://nces.ed.gov/nationsreportcard/tdw/analysis/2000_2001/infer_multiplecompare_fdr.asp. Unlike other multiple comparison procedures that control the family-wise error rate (i.e., the probability of making even one false rejection in the set of comparisons), the FDR procedure controls the expected proportion of falsely rejected hypotheses. Furthermore, the FDR procedure used in NAEP is considered appropriately less conserva-tive than family-wise procedures for large families of com-parisons (Williams, Jones, and Tukey 1994). Therefore, the FDR procedure is more suitable for multiple comparisons in NAEP than are other procedures.

NAEP employs a number of rules to determine the num-ber of comparisons conducted, which in most cases is sim-ply the number of possible statistical tests. However, there are two exceptions where the FDR is not applied: when comparing multiple years and when comparing multiple jurisdictions to the nation, neither the number of years nor the number of jurisdictions counts toward the number of comparisons.

Cautions in interpretation It is possible to examine NAEP performance results for groups of students defined by various background fac-tors measured by NAEP. However, a relationship that exists between achievement and another variable does not reveal its underlying cause, which may be influenced by a number of other variables. The results are most use-ful when they are considered in combination with other knowledge about the student population and the educa-tional system, such as trends in instruction, changes in the school-age population, and societal demands and expecta-tions, among others.

Appe

ndix

A: T

echn

ical

Not

es

http://nces.ed.gov/nationsreportcard/tdw/analysis/2000_2001/infer_multiplecompare_fdr.asp

76

Appendix B: Supplemental Tables

Table B-1. Administration of NAEP national and state mathematics assessments, by grade: Various years, 1990–2009

1990 1992 1996 2000 2003 2005 2007 2009 Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State

4th grade U U U U U U U U U U U U U U U8th grade U U U U U U U U U U U U U U U USOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress, Various years, 1990–2009 Mathematics Assessments.

Table B-2. Average national mathematics scale scores for all public school students at grades 4 and 8, by gender and eligibility for the National School Lunch Program: Various years, 1990–2009

1990n 1992n 1996 2000 2003 2005 2007 2009All Students

Grade 4 212* 219* 222* 224* 234* 237* 239 239Grade 8 262* 267* 269* 272* 276* 278* 280* 282

Student Gender Grade 4

Male 212* 220* 222* 225* 235* 238* 240 240Female 211* 218* 222* 223* 233* 236* 238 238

Grade 8 Male 262* 266* 270* 273* 277* 278* 281* 283Female 261* 267* 268* 271* 275* 277* 279* 281

Student Eligibility for National School Lunch Program

Grade 4 Not Eligible — — 232* 235* 244* 248* 249* 250Eligible — — 207* 208* 222* 225* 227 228

Grade 8 Not Eligible — — 277* 283* 287* 288* 291* 293Eligible — — 250* 253* 258* 261* 265* 266

n Accommodations were not permitted for this assessment.— Not available. Data were not collected prior to 1996.* Significantly different (p<.05) from 2009.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), Various years, 1990–2009 Mathematics Assessments.

77

Table B-3. Administration of NAEP national and state reading assessments, by grade: Various years, 1992–2009

1992 1994 1998 2000 2002 2003 2005 2007 2009 Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State Nat’l State4th grade U U U U U U U U U U U U U U U U U8th grade U U U U U U U U U U U U U USOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress, Various years, 1992–2009 Reading Assessments.

Table B-4. Average national mathematics scale scores for all public school students at grades 4 and 8, by gender and eligibility for the National School Lunch Program: Various years, 1990–2009

1992n 1994n 1998 2000 2002 2003 2005 2007 2009All Students

Grade 4 215* 212* 213* 211* 217* 216* 217* 220 220Grade 8 258* 257* 261 — 263 261* 260* 261* 262

Student Gender Grade 4

Male 211* 207* 210* 206* 214* 213* 214* 216 216Female 219* 218* 215* 217* 220* 220* 220* 223 223

Grade 8 Male 251* 250* 253* — 258 256* 255* 256* 258Female 264 265* 268 — 267 267 266* 266* 267

Student Eligibility for National School Lunch Program

Grade 4 Not Eligible — — 226* 226* 229* 229* 230* 232 232Eligible — — 195* 192* 202* 201* 203* 205* 206

Grade 8 Not Eligible — — 268* — 271* 271* 270* 271* 273Eligible — — 245* — 249 246* 247* 247* 249

n Accommodations were not permitted for this assessment.— Not available. Data were not collected prior to 1996 or at grade 8 in 2000.* Significantly different (p<.05) from 2009.SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), Various years, 1992–2009 Reading Assessments.

78

Table B-5. Percentages of public school students in NAEP mathematics and reading classified as English language learners, by subject, grade, race/ethnicity, and jurisdiction: 2009

Mathematics Reading Grade 4 Grade 8 Grade 4 Grade 8

Hispanic White Hispanic White Hispanic White Hispanic WhiteNational Public 37 2 21 # 35 1 20 #

Alabama 46 # 41 # 44 # 35 #Alaska 18 2 19 1 14 1 19 1Arizona 29 1 12 # 29 # 12 #Arkansas 59 # 45 # 62 # 42 #California 49 4 32 1 49 4 32 1Colorado 32 # 21 1 33 1 19 #Connecticut 23 1 16 1 18 1 10 #Delaware 23 # 13 # 24 # 11 #District of Columbia 48 5 21 ‡ 43 3 28 ‡DoDEA1 13 2 10 2 9 2 — —Florida 24 1 13 # 17 1 10 #Georgia 30 # 14 # 25 # 14 #Hawaii 12 # 6 # 13 1 6 1Idaho 28 1 21 # 29 # 23 #Illinois 26 1 10 1 26 1 11 #Indiana 45 # 31 1 47 # 29 #Iowa 41 # 20 # 41 1 24 #Kansas 53 # 35 # 49 # 33 #Kentucky 33 # 23 # 34 # 10 #Louisiana 42 # ‡ # 47 # ‡ #Maine ‡ # ‡ # ‡ # ‡ #Maryland 32 # 16 # 23 1 5 #Massachusetts 25 1 10 # 25 1 10 #Michigan 19 2 25 1 24 2 22 1Minnesota 42 1 24 # 44 # 32 #Mississippi ‡ # ‡ # 28 # ‡ #Missouri 25 # 10 # 25 # 6 #Montana 2 1 3 # 4 1 ‡ #Nebraska 34 1 18 # 31 1 15 #Nevada 47 1 19 # 45 1 17 #New Hampshire 29 1 17 # 25 1 6 #New Jersey 11 # 8 # 5 # 3 #New Mexico 23 1 14 # 17 1 12 #New York 27 1 17 1 24 1 11 #North Carolina 43 # 40 # 43 # 39 #North Dakota ‡ # ‡ # ‡ # ‡ 1Ohio 29 # 18 # 35 # 13 #Oklahoma 35 # 25 # 33 # 21 #Oregon 51 1 30 1 51 1 27 1Pennsylvania 19 # 16 # 19 # 15 #Rhode Island 23 1 10 # 21 # 8 #South Carolina 66 1 49 # 61 1 49 #South Dakota 18 1 7 # 11 1 ‡ 1Tennessee 36 # 14 # 39 # 15 #Texas 37 1 12 # 31 1 13 #Utah 41 # 26 # 37 # 22 #Vermont ‡ 1 ‡ 1 ‡ 1 ‡ #Virginia 49 1 21 # 50 1 27 #Washington 37 2 14 1 35 1 14 #West Virginia ‡ # ‡ # ‡ # ‡ #Wisconsin 46 1 30 # 44 # 23 #Wyoming 14 # 11 # 17 # 7 #— Not available. # Rounds to zero.‡ Reporting standards not met. 1 Department of Defense Education Activity (overseas and domestic schools).SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Mathematics and Reading Assessments.

79

Table B-6. Percentages of public school students in NAEP mathematics and reading classified as eligible for a free or reduced-price school lunch, by subject, grade, race/ethnicity, and jurisdiction: 2009

Mathematics Reading Grade 4 Grade 8 Grade 4 Grade 8

Hispanic White Hispanic White Hispanic White Hispanic WhiteNational Public 77 29 72 25 74 29 72 24

Alabama 84 39 70 33 85 39 85 33Alaska 59 28 45 21 48 28 50 21Arizona 75 28 71 22 75 29 71 21Arkansas 91 46 84 41 85 46 85 40California 75 23 73 22 74 24 73 20Colorado 70 21 68 18 72 20 68 16Connecticut 73 13 70 11 72 12 67 12Delaware 74 25 69 22 74 24 72 20District of Columbia 75 7 80 ‡ 79 5 81 ‡DoDEA1 # # # # # # — —Florida 72 35 68 27 70 36 65 27Georgia 90 33 79 29 82 33 75 28Hawaii 45 31 43 28 41 32 47 25Idaho 76 37 73 29 75 36 74 29Illinois 79 22 71 17 78 23 71 17Indiana 78 37 75 28 79 36 75 28Iowa 81 30 73 27 78 30 76 27Kansas 85 37 84 31 84 36 83 31Kentucky 81 47 8 43 77 46 82 44Louisiana 74 49 ‡ 42 77 50 ‡ 43Maine ‡ 39 ‡ 33 # 38 ‡ 33Maryland 67 18 60 14 65 19 57 12Massachusetts 83 18 73 16 80 16 77 17Michigan 73 33 53 29 70 34 69 28Minnesota 62 21 64 17 69 21 68 16Mississippi ‡ 48 ‡ 43 86 46 ‡ 43Missouri 68 36 52 30 75 35 66 30Montana 50 34 48 28 62 34 ‡ 29Nebraska 80 30 78 27 77 31 80 25Nevada 58 23 53 19 59 24 51 20New Hampshire 61 20 56 19 58 20 57 18New Jersey 72 10 63 10 70 11 61 10New Mexico 81 40 77 33 79 38 77 32New York 83 26 75 22 83 25 75 20North Carolina 82 28 78 24 83 28 75 25North Dakota ‡ 27 ‡ 23 ‡ 27 ‡ 22Ohio 73 29 58 26 63 29 65 26Oklahoma 77 45 78 37 78 42 78 38Oregon 85 35 81 31 85 36 79 31Pennsylvania 79 25 81 23 78 25 78 23Rhode Island 84 24 80 22 83 24 80 22South Carolina 76 36 68 32 84 36 77 29South Dakota 67 27 63 24 68 26 ‡ 23Tennessee 72 40 70 32 80 40 68 32Texas 81 23 75 21 78 26 75 22Utah 75 25 68 18 78 24 71 18Vermont ‡ 34 ‡ 28 ‡ 34 ‡ 28Virginia 64 19 61 19 58 19 56 18Washington 83 33 77 27 52 32 76 26West Virginia ‡ 56 ‡ 51 ‡ 56 ‡ 50Wisconsin 73 27 67 21 78 27 69 20Wyoming 64 29 55 24 59 30 54 24— Not available. # Rounds to zero.‡ Reporting standards not met. 1 Department of Defense Education Activity (overseas and domestic schools).SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2009 Mathematics and Reading Assessments.

80

This report was prepared by the NAEP Education Statistics Services Institute (NAEP ESSI) of the American Institutes for Research (AIR) under contract with the National Center for Education Statistics (NCES), part of the Institute of Education Sciences of the U.S. Department of Education.

The authors are grateful to everyone who contributed to the design, development, review, and production of this report.

Acknowledgments

achievement gaps: how hispanic and white students in ... · achievement gap between hispanic and...

Documents